ProModel User Guide-3

Download as pdf or txt
Download as pdf or txt
You are on page 1of 101

184 Chapter 5:

Cost

Per With this pull-down menu, you can set the cost applies to the resource only when it returns
time units for the Operation Rate. Time units may to the entity.
be in seconds, minutes, hours, or days as shown
here.

Entities

Resources

Initial Cost This field allows you to define the


initial entity cost for an entity which enters the
system through a scheduled arrival, a CREATE
statement, or an ORDER statement. ProModel
Regular Rate This field specifies the cost per accepts expressions in this field and evaluates
unit of time for a resource used in the model. You them at translation.
can use expressions in this field (evaluated at
translation) to set the rate or change it using Set- Please note
Rate. For more information on the SetRate opera-
tion statement, see “SetRate” on page 556. When you implicitly create new entities through a
Per This pull-down menu allows you to set the
ROUTE statement, ProModel does not add an
time units for the Regular Rate. Times may be in initial cost to the entity. To add an initial cost,
seconds, minutes, hours, or days as shown here. use the INCENTCOST statement. See “IncEnt-
Cost” on page 499 for more information.

Building a Model with Costing


Cost Per Use This field allows you to define the
actual dollar cost accrued each time you use the When you build a model using the costing fea-
resource (i.e., the minimum usage cost). The cost ture, you must first define the locations,
per use updates when you obtain the resource and resources, and entities used in the model. Once
ProModel accepts expressions in this field (eval- you define these model components, you may
uated at translation). assign costing information to them through the
Cost option in the Build menu. To collect costing
Please note information about your model, uncheck the dis-
able costing box from the simulation options dia-
log of the simulation menu. By default, ProModel
Since ProModel counts a preemption as a use, if
you preempt a resource from an entity, the usage
ProModel 185
User Guide

disables costing and sets all defaults to zero. See both entities. If the preempting entity
“Enable or Disable Costing” on page 187. obtains a resource, the preempted entity
will not record the resource costs during
the preemption period.
•If an entity is at a location when a preemp-
tive downtime occurs, the entity records
the downtime as part of its operational
costs. This applies to all types of location
downtimes, including shifts. The location
records the cost of the preempted entity
while it remains at the location.
•If an entity’s resource has a downtime which
requires the use of another resource, the
entity will not record the second
resource’s cost. However, the location will
record the extra resource’s cost.

Join/Load
•Joined entities add their costs to their base
entities, but not their time statistics.
•Loaded entities do not add their costs or time
statistics to their base entities.
Please note •When an UNLOAD occurs, ProModel
divides all costs accrued by a loaded entity
The following scenarios assume you defined costs among the unloaded entities. ProModel
for all model components. adds all other entity statistics calculated
during the loaded period to each of the
unloaded entities.
•Entities leaving the system loaded onto other
entities do NOT report their individual
Preemption/Downtime costs, but do report all other statistics. To
•If you preempt an entity’s resource, an addi- get the cost of each entity, you must
tional cost per use will apply once you re- unload the entities before they exit.
acquire the resource. While waiting for the
resource to return, the entity does not
record operation or resource costs. Combine/Group
•If an entity preempts another entity, the pre- •Combined entities add their costs to the
empted entity continues to record opera- resultant entity, but not their time statis-
tion time during the entire preemption tics. The resultant entity begins with fresh
period. While the preempting and pre- time statistics.
empted entities are simultaneously at a •Grouped entities do not add their costs or
location, the location records the cost for statistics to the group shell (a temporary
186 Chapter 5:
Cost

entity representing grouped entities that Costing Output Statistics


starts with cost and time statistics of zero).
•When an UNGROUP occurs, ProModel ProModel collects costing statistics only if you
divides all costs accrued by a grouped uncheck the Disable Cost Statistics option in the
entity among the ungrouped entities. Pro- Simulation Options menu (see “Enable or Dis-
Model copies all other entity statistics cal- able Costing” on page 187). Included in the Gen-
culated during the grouped period to each eral Statistics Report, ProModel calculates
of the ungrouped entities. costing statistics.
•Entities leaving the system grouped with
other entities do NOT report their individ- Locations
ual costs, but do report all other statistics. •Operational Cost = (Active Operation Time
To get the cost of each entity, you must * Rate) + (Any IncLocCost)
ungroup the entities before they exit. •% Operational Cost refers to the location’s
percentage of the sum of all operation
costs
Special Cost Handling •Resource Cost = (Utilization * Rate) +
•As soon as you acquire a resource, it begins (Times Used * Cost per use)
to accrue cost.
•Unless obtained in the move logic, Pro- Please note
Model charges the “Cost per use” for a
resource to the location that obtained it. For Resource Cost, Utilization and Times Used
Resources obtained in the move logic do refer to the utilization of a resource while at a
not charge a “per use” cost to any location. location. This applies only to resource use
•ProModel does not charge any resource time through operation logic.
used during move logic to any location.
•ProModel adds initial entity costs defined in
the cost module only as entity costs, not
location costs. •% Resource Cost refers to the location’s per-
•If a location uses a resource during a down- centage of the sum of all resource costs
time, the location accrues that resource’s •Total Cost = (Operation Cost + Resource
cost. Cost)
•The USE statement counts as operation and •% Total Cost refers to location’s percentage
resource cost. of the sum of all location costs
•When you CREATE a new entity, it begins
with new time statistics and an initial cost. Resources
•If you RENAME an entity, previous time •NonUse Cost = (1-% Utilization) * Sched-
statistics and costs continue with the uled Time * Rate
entity. •% NonUse Cost refers to the resource’s per-
•The SPLIT AS statement divides the cost of centage of the sum of all nonuse costs
the original entity between all entities. •Usage Cost = (% Utilization * Scheduled
Each new entity begins with new time sta- Time * Rate) + (Times Used * Cost per
tistics. use)
ProModel 187
User Guide

•% Usage Cost refers to the resource’s per- default time units specified in the General Infor-
centage of the sum of all resource usage mation dialog.
costs
In the above calculations, the rate defined (per
•Total Cost = Usage Cost + NonUse Cost
day, hour, minute, and second) converts to the
•% Total Cost refers to the resource’s per- default time units specified in the General Infor-
centage of the sum of all resource costs mation dialog.

Entities
Please note
• Explicit Exits The number of entities that
have explicitly exited. Whenever an entity
exits the system, it is an explicit exit ProModel does not allow you to generate a Cost-
except in the following cases: ing Graph. However, if you set a variable equal
- When an entity JOINS or COMBINES to GetCost (e.g., Var1=GetCost), you can gener-
with another entity, it implicitly exits the ate a time series graph to track changing entity
system, and is reported as an exit in the costs. See “GetCost()” on page 488 for more
Entity Acitvity report. However, for information.
costing purposes, the entity did not
explicitly exit, but its costing informa-
tion was added to the entity it was
JOINED or COMBINED with. Enable or Disable Costing
- When an entity LOADS or GROUPS
To enable the costing feature, be sure that the
with another entity, and the entire
Disable Cost option in the Simulation Options
LOADED or GROUPED entity exits the
dialog is not checked.
system, the original entity implicitly
exits the system, and is reported as an
exit in the Entity Acitvity report. How-
ever, for costing purposes, the original
entity did not explicitly exit, but its cost-
ing information was added to the entire
load or group.
• Total Cost Dollars Total Cost = cumula-
tive entity cost, or the sum of costs
incurred on all locations the entity passed
through + the sum of all costs incurred by
use of resource + initial cost + any IncEnt-
Cost
• % Total Cost % Total Cost refers to the
entity’s percentage of sum of all entity
costs
In the above calculations, the rate defined (per
day, hour, minute, and second) converts to the
188 Chapter 5:
Tanks

Tanks Please note


Tanks are simply locations to which ProModel When you create the first tank in your model,
associates a level instead of an entity routing. (As ProModel will display a dialog that allows you to
a result, the units and rules fields do not apply to automatically import various subroutines,
tanks.) Using tanks, you can model the continu- arrays, macros, and library graphics specific to
ous flow of liquids and other substances into and tanks. If you do not wish to include these new
out of tanks or similar vessels. Also, when com- items, you may cancel the action.
bined with discrete-event simulation, ProModel’s
continuous modeling capability makes it possible
to model the exchange between continuous mate-
rial and discrete entities (e.g., when you place liq- 3. Enter a capacity (1 to 999999) for the tank
uid into a container). Other uses include in the location capacity field.
modeling high-rate, discrete part manufacturing
4. Define and reference any necessary tank
systems. control subroutines.

The Tank Submodel


In order to function properly, all tank models
How to edit a tank or a gauge
should include the tank submodel (TANK-
SUB.MOD). The tank submodel contains impor-
1. Double click on the tank or gauge (or right
tant subroutines and data elements (arrays and
click and select Edit Graphic).
macros) used to simplify tank modeling. Each of
these subroutines and data elements has a 2. From the dialog that appears, make the
“Tank_” prefix to help identify it and to prevent appropriate changes.
any accidental name duplication. 3. Click OK.

Please note

All user-defined model elements should begin


with something other than “Tank_”.

How to define a tank

1. Select the gauge/tank symbol from the


Location Graphics window.
2. Click on the layout window where you wish
to place the tank and select Create Tank
Location from the menu that appears. Pro-
Model places the tank on the layout.
ProModel 189
User Guide

How to change between a tank the Subroutine module. To call these subroutines
and a gauge and operate them independently in the model, use
the ACTIVATE statement. For examples of how
to use these subroutines, see the discussion at the
• Double click on the tank or gauge and
check or uncheck the tank option. end of this section. To model tanks effectively,
you must understand the following concepts.

or... Tank Levels


ProModel records tank levels in a pre-defined
array called Tank_Level where each element of
• Right click on the tank or gauge and select
Change Tank to Gauge or Change Gauge to the array corresponds to each tank location in the
Tank. location list. For example, the level of TankA is
the value of Tank_Level [TankA]. If TankA were
the third location in the location list, you could
also reference the level of TankA with
In addition to defining a tank graphic, you may Tank_Level[3]. For best results, you should con-
add labels and other figures to a tank. For exam- trol all tank levels using only the pre-defined tank
ple, you can add a counter to digitally display the filling and emptying subroutines rather than
fill level of the tank (ProModel rounds the value change the Tank_Level array values directly.
displayed to the nearest integer). The following This will prevent overfilling or overdrawing and
are examples of how you can use tanks in Pro- will accurately gather statistics for each tank. For
Model. example, calling Tank_Fill (TankA, 500, 30, 0)
automatically fills TankA to 500 units at a rate of
30 units per minute. The 0 signifies that the tank
will not accept excess material and, as a result, an
error message will occur if the tank reaches
capacity before the specified amount fills into the
tank.

Basic Concepts
Since tanks do not process discrete entities, you
may not define routings to or from tanks. To con-
trol a tank level, ProModel provides predefined
subroutines that fill, empty, and transfer tank
contents. To monitor tank levels and initiate
flows, you must define control subroutines using
190 Chapter 5:
Tanks

The Flow Time Step flow. The units of flow, however, may change
when you move material from one tank to
To model continuous flow, ProModel uses a another (e.g., pounds of dry material may transfer
Tank_TimeStep macro. This macro is the time into a tank containing gallons of liquid).
step used when filling/emptying tanks and is an
RTI (run-time interface) parameter. Initially, Pro- To specify a variable rate of flow that changes
Model sets this value to .2 minutes. If you wish to dynamically with each time step, pass a value of
use a different value for the time step, you may 0. This signals the subroutine to call the
change it temporarily (for a particular model) Tank_Rate subroutine with each time step. To
through the Simulation/Parameters menu option, return the desired rate value for each time step
or permanently by changing the macro itself. The when you use a variable rate, you must modify
larger the time step, the longer the interval the Tank_Rate subroutine appropriately.
between filling and emptying (which speeds up
the simulation). For example, suppose you set the Tank States
time step to .1 minutes. If a tank empties at a rate
of 60 gpm, the simulation would actually empty Like other model elements, tanks use states to
the tank by a discrete amount of 6 gallons every test and track statistics. ProModel automatically
.1 minutes. When filling or emptying a tank, if sets these states when you use the predefined tank
the remaining quantity doesn't require the full subroutines to control the tank. The following are
time step, ProModel reduces the time step using a defined states:
linear interpolation. Tank_Idle The tank is empty and not in use. Set
automatically when a tank empties and at the end
Please note of a Tank_DoPrep or Tank_GoDown subroutine.
Tank_Operation The tank is active (e.g., mix-
The only adverse effect of using a large time step ing, reacting, heating). Set automatically when
is that any WAIT UNTIL statement or other test the model calls the Tank_DoOperation subrou-
based on the Tank_Level array may be off by as tine.
much as the flow amount for the time step. For
example, if the time step is .5 minutes and the rate Tank_Setup The tank is cleaning or preparing
of flow is 60 gpm, the level will change in 30 gal- for future use. Set automatically whenever you
lon increments. This means that the tank will not call the Tank_Prep subroutine.
satisfy the statement “WAIT UNTIL Tank_Filling The tank is filling. Set automati-
Tank_Level[TankA]>=31” until the level cally whenever you fill the tank.
reaches 60.
Tank_Emptying The tank is emptying. Set auto-
matically whenever you empty the tank.
Tank_Blocked The tank is full and ready to
Rate of Flow transfer. Set automatically when the tank fills to
To use flow rates properly, you must define all capacity.
rates in terms of units (i.e., gallons, pounds) per Tank_Down The tank is down. Set automati-
time unit defined in the General Information dia- cally whenever you call Tank_GoDown.
log. Whenever you call one of the empty, fill, or
transfer subroutines, you must specify the rate of
ProModel 191
User Guide

Tank_ScheduledDown Similar to the SKIP example


Tank_Down state, except statistics are not col-
lected. Tank_GoDownSched (<TankID>, DTLeft())
While ProModel sets these states automatically, SKIP
you may change the state of the tank at any time
by calling the Tank_SetState subroutine. Pro-
Model records statistics for these states in the
ProModel temporarily suspends tank flow while
output report under Locations. Since a tank may
a tank is down or off shift.
fill and empty simultaneously, the output report
combines Tank_Filling with Tank_Emptying and
reports it all as waiting time. Tank Logic Builder
An expanded capability within ProModel, the
Over Filling/Emptying Tanks tank logic builder provides you with what you
When using the predefined subroutines to fill, need to model complex tank and fluid system
empty, or transfer from one tank to another, you operations. The logic builder contains all avail-
may accidentally attempt to over fill or over able tank subroutines and provides you with a
empty a tank. To prevent these situations, you description of the components required to use
have the option to terminate the fill/empty sub- each subroutine.
routine or suspend further filling/emptying until
the tank reaches a resume level. If you terminate
the subroutine, ProModel temporarily stores the
un-filled or un-emptied quantity for immediate
access in the global variable, Tank_QtyLeft.

Tank Downtimes
For Tanks, you must define downtimes and shifts
in a special way. First, you may define only clock
downtimes for tanks. Second, when defining a
clock downtime for a tank, use the
Tank_GoDown subroutine (page 201) in the
Downtime Logic field instead of just a WAIT
statement. This sets the state of the tank to
Tank_Down and gathers the appropriate statis-
tics. Third, when defining a shift for a tank, you
should call the Tank_GoDownSched subroutine
in the off-shift logic using the DTLeft() function
as the time parameter. A SKIP statement should
follow this function as shown next.
192 Chapter 5:
Tanks

Please note

The subroutine logic is not accessible until you


define your first tank location—when you define
the tank location, ProModel loads the tank sub-
model.

How to access the Logic Builder:

• Click the right mouse button in the logic


window or expression edit field. Or click the
Build button on the logic window’s toolbar.

For more information about the Logic Builder,


see “Logic Builder” on page 293.
ProModel 193
User Guide

Pre-defined Tank Subrou- TANK_FILL (HoldingTank, 2000, 75, 1500)

tines
Tank_Cap .............................................. page 200 Description
Tank_Dec............................................... page 198 Fills a tank using a specific quantity and rate. The
Tank_DoOperation ............................... page 201 default tank state sets to Tank_Filling, then to
Tank_Blocked if the tank becomes full.
Tank_DoPrep ......................................... page 203
Use Tank_Fill when the source of the material is
Tank_Empty ........................................... page 194 not another tank, but an arriving entity or a
Tank_FallTrigger .................................... page 200 source that is not part of the model.

Tank_Fill ................................................. page 193


Components
Tank_FreeCap ....................................... page 200
<Tank ID>
Tank_GoDown....................................... page 201
The tank name or location index number.
Tank_GoDownSched............................ page 202
<Fill Quantity>
Tank_Inc ................................................ page 198
The number of units (gallons, pounds) to fill into the
Tank_Rate .............................................. page 207 tank. To fill the tank to capacity, enter
Tank_Cap(<Tank ID>).
Tank_RiseTrigger ................................... page 199
<Fill Rate>
Tank_SelectInput .................................. page 206
The rate in units (gallons, pounds) per time unit
Tank_SelectOutput ............................... page 204 defined in the General Information dialog. To
instantly increase the level of a tank, use the Tank_Inc
Tank_SetLevel ....................................... page 197
subroutine. To initialize the level of a tank (e.g., at the
Tank_SetState ........................................ page 204 start of the simulation), use the Tank_SetLevel subrou-
tine. To use a dynamically calculated rate in the
Tank_Transfer ......................................... page 195 Tank_Rate subroutine, enter 0.
Tank_TransferDownTo ........................... page 196 <Resume level>
Tank_TransferUpTo ................................ page 195 If the tank level reaches capacity before you add the
specified quantity, the tank must drop to the resume
Tank_UpdateStats ................................. page 206 level before it can continue filling. To terminate filling
if the tank reaches capacity, enter Tank_Stop as the
resume level. A value of 0 causes an error to occur if
Tank_Fill the tank becomes full before reaching the fill quantity.

Syntax samples
Example
TANK_FILL (<Tank ID>, <Fill Quantity>, <Fill
Rate>, <Resume Level>)
194 Chapter 5:
Pre-defined Tank Subroutines

A tanker arrives and fills a storage tank by the The tank name or location index number.
quantity stored in the tanker’s attribute,
<Empty Quantity>
Load_Qty. The rate of fill is 80 gpm and, if the
tank fills to capacity before the tanker dis- The number of units (gallons, pounds) to empty. To
charges the entire quantity, the level of the empty a tank completely of its current contents, enter
storage tank must drop to 12,000 gallons Tank_Level [<Tank ID>].
before it resumes filling. To represent this, <Empty Rate>
enter the following statement in the opera-
tion logic for the tanker at the unloading sta- The rate in units (gallons, pounds) per time unit
tion. defined in the General Information dialog. To instantly
decrease the level of a tank, use the Tank_Dec subrou-
Tank_Fill(StorageTank, Load_Qty, 80, 12000) tine. To specify a dynamically calculated rate using the
Tank_Rate subroutine, enter 0.
<Resume level>
See Also If the tank level drops to 0 before you empty the speci-
fied quantity, the tank must rise to the resume level
“Filling from an Entity” on page 210 and “Initial-
before continuing to empty. To terminate emptying if
izing and Replenishing Supply Tanks” on the level ever drops to 0, enter Tank_Stop. A value of 0
page 210. causes an error to occur if the tank becomes empty
before removing the specified quantity.
Tank_Empty
Example
Syntax samples When a chemical tank, ChemTank, is full
(state is Tank_Blocked), workers pump its con-
TANK_EMPTY (<Tank ID>, <Empty Quantity>, tents into a rail car at a rate of 60 gpm for
<Empty Rate>, <Resume Level>) transportation to another facility. Since rail
TANK_EMPTY (TankB, 2000, 40, 0) cars are always available and the delivery
activity is not of interest, it is not necessary to
model the rail cars explicitly. Instead, activate
a subroutine in the initialization logic with the
following statement:
Description
Tank_Loop //logic repeats continuously
Empties a tank by a specified quantity and rate.
The state is set to Tank_Emptying, then to {
Tank_Idle if the tank becomes empty. WAIT UNTIL
Tank_State[ChemTank]=Tank_Blocked
Use Tank_Empty when the destination is not
another tank, but an arriving entity or a source Tank_Fill(ChemTank, Tank_Level[ChemTank],
that is not part of the model. 60, 0)
}

Components

<Tank ID>
ProModel 195
User Guide

See Also The rate in units (gallons, pounds) per time unit
defined out of the FROM tank. To use a dynamically
“Emptying to an Entity” on page 211. calculated rate in the Tank_Rate subroutine, enter 0.
<TO Rate>
Tank_Transfer The rate in units (gallons, pounds) per time unit
defined in the General Information dialog into the TO
Syntax samples tank. Use 0 if same as the FROM rate. (The TO rate is
automatically the same as the FROM rate if you add 0
as the FROM rate.)
TANK_TRANSFER (<FROM Tank ID>, <TO Tank
ID>, <Transfer Quantity>, <FROM Rate>, <TO <Resume level>
Rate>, <Resume Level>) If the TO tank reaches capacity before the specified
TANK_TRANSFER (Tank1, Tank2, 2000, 100, 0, quantity transfers, the TO tank must drop to the resume
0) level before continuing with the transfer. To terminate
transferring when the TO tank reaches capacity, enter
Tank_Stop. A value of 0 causes an error to occur if the
tank becomes empty before transferring the specified
Description quantity.

Transfers a specified quantity from one tank to


Example
another. ProModel sets the state of the FROM
tank to Tank_Emptying and the TO tank to
When a mixing tank is ready to mix a new
Tank_Filling. If the FROM tank becomes empty,
batch of material, 10,000 gallons of water
its state becomes Tank_Idle. If the TO tank
must first transfer from a water supply tank at
becomes full, its state becomes Tank_Blocked. a rate of 100 gpm. The following logic repre-
Otherwise, the states remain unchanged. sents this action:
Use Tank_Transfer when you want to transfer a Tank_Transfer(WaterTank, MixingTank, 10000,
specific quantity from one tank to another. 100, 0, 0)

Components
See Also
<FROM Tank ID>
“Tank Transfers” on page 212.
The name or location index number of the FROM tank.
<TO Tank ID>
The name or location index number of the TO tank.
Tank_TransferUpTo
<Transfer Quantity>
Syntax samples
The number of units (gallons, pounds) to transfer. To
transfer the entire contents of a tank, enter Tank_Level
[<FROM Tank ID>]. TANK_TRANSFERUPTO (<FROM Tank ID>, <TO
Tank ID>, <TO Level >, <FROM Rate>, <TO
<FROM Rate> Rate>)
196 Chapter 5:
Pre-defined Tank Subroutines

TANK_TRANSFERUPTO (Tank1, Tank2, 8500, 75, Example


0)
An in-process tank supplies several down-
stream tanks and must maintain a maximum
level of 20,000 gallons. Whenever the in-pro-
Description cess tank drops below 5,000 gallons, a supply
Similar to Tank_Transfer except that tank refills the tank at a rate of 100 gpm. To
Tank_TransferUpTo does NOT terminate a trans- model this, define an activated subroutine for
the supply tank using the following logic:
fer based on the transferred quantity, rather when
the TO tank level rises to a certain point. If the Tank_Loop //logic repeats continuously
tank empties before reaching the TO level, Pro- {
Model suspends the transfer until capacity
WAIT UNTIL Tank_Level[InProcessTank]<=5000
becomes available.
Tank_TransferUpTo(SupplyTank, InProc-
Use Tank_TransferUpTo when you want to raise essTank, 20000, 100, 0)
the level of a tank to a certain value but are not
certain of the quantity needed to reach that level }
(e.g., the tank is draining at the same time you are
trying to fill it).
Tank_TransferDownTo
Components

<FROM Tank ID> Syntax samples


The name or location index number of the FROM tank.
TANK_TRANSFERDOWNTO (<FROM Tank ID>,
<TO Tank ID> <TO Tank ID>, <TO Level >, <FROM Rate>, <TO
The name or location index number of the TO tank. Rate>)

<TO Level> TANK_TRANSFERDOWNTO (Tank1, Tank2, 1000,


80, 0)
Transfer until the TO tank reaches this level.
<FROM Rate>
The rate in units (gallons, pounds) per time unit
Description
defined in the General Information dialog out of the
FROM tank. To use a dynamically calculated rate in Similar to Tank_Transfer except that
the Tank_Rate subroutine, enter 0. Tank_TransferDownTo terminates the transfer
<TO Rate> when the FROM tank level lowers to a desig-
nated level instead of lowering by a specific
The rate in units (gallons, pounds) per time unit
quantity. If the TO tank becomes full, ProModel
defined in the General Information dialog into the TO
tank. Use 0 if the TO rate is the same as the FROM suspends the transfer until capacity becomes
rate. available.
Use Tank_TransferDownTo when you want to
lower the level of a tank to a specific value but
ProModel 197
User Guide

you are not certain how much to empty in order Tank_TransferDownTo(TankA, TankB, 200, 50,
to drop to that level (e.g., the tank may fill at the 0)
same time it empties). }

Components

<FROM Tank ID> See Also


The name or location index number of the FROM tank. “Split Transfers” on page 214.
<TO Tank ID>
The name or location index number of the TO tank. Tank_SetLevel
<TO Level>
Transfer until the FROM tank drops to this level.
Syntax samples
<FROM Rate> TANK_SETLEVEL (<Tank ID>, <Quantity>)
The rate in units (gallons, pounds) per time unit TANK_SETLEVEL (TankA, 1500)
defined in the General Information dialog out of the
FROM tank. To use a dynamically calculated rate in
the Tank_Rate subroutine, enter 0.
<TO Rate> Description
The rate in units (gallons, pounds) per time unit Instantly sets the level of a tank to a specified
defined in the General Information dialog into the TO quantity. If the quantity is negative or larger than
tank. Use 0 if the TO rate is the same as the FROM
the tank capacity, an error occurs. The tank state
rate.
sets to Tank_Blocked if you set the tank level to
the tank capacity and to Tank_Idle if you set the
tank level to 0. Otherwise, the state remains
Example unchanged.

An in-process tank, TankA, supplies TankB at a Use Tank_SetLevel when you want to initialize a
rate of 50 gpm. TankA must maintain a mini- tank to a specific level.
mum level of 200 gallons to insure against
pump cavitation. When TankA’s level drops Components
to 200 gallons, the tank stops pumping to
TankB until the level of TankA rises above 200
<Tank ID>
gallons. To model this scenario, enter the fol-
lowing logic in the subroutine controlling the The tank name or location index number.
flow from TankA to TankB:
<Quantity>
Tank_Loop //logic repeats continuously
The level at which to set the tank (number of gallons,
{ pounds). To completely fill the tank, enter
Tank_Cap(<Tank Name>).
WAIT UNTIL Tank_Level[TankA]>200
198 Chapter 5:
Pre-defined Tank Subroutines

Example The number of units by which to increment the con-


tents of the tank (gallons, pounds).
When you begin a simulation, you wish to set
the initial level of a supply tank, TankX, to
10,000 gallons. To model this, enter the follow- Example
ing statement in the initialization logic for the
model. Trucks deliver pellets to a holding bin twice a
Tank_SetLevel(TankX, 10000) day. When a truck arrives at the drop-off sta-
tion, it dumps the entire 5,000 lb load in only
2.5 minutes. To model this, define the follow-
ing operation logic for the truck at the drop-
See Also off station:
WAIT 2.5 MIN
“Initializing and Replenishing Supply Tanks” on
page 210. Tank_Inc(HoldingBin, 5000)

Tank_Inc
Tank_Dec
Syntax samples
Syntax samples
TANK_INC (<Tank ID>, <Quantity>)
TANK_INC (StorageTank, 5000) TANK_DEC (<Tank ID>, <Quantity>)
TANK_DEC (SupplyTankB, 1000)

Description
Instantly increases the level of a tank by a speci- Description
fied quantity. If the tank has insufficient capacity, Instantly decreases the level of a tank by a speci-
the level increases as capacity becomes available. fied quantity. If the tank has insufficient quantity,
ProModel sets the tank state to Tank_Blocked if it empties as material becomes available. Pro-
the level increases to the tank capacity, otherwise Model sets the tank state to Tank_Idle if you
the state remains unchanged. decrease the level to 0. Otherwise the state
Use Tank_Inc to instantly add a specific quantity remains unchanged.
to a tank. Use Tank_Dec to instantly remove a specific
quantity from a tank.
Components
Components
<Tank ID>
The tank name or location index number. <Tank ID>
<Quantity> The tank name or location index number.
ProModel 199
User Guide

<Quantity> Components
The number of units by which to decrement the con-
tents of the tank (gallons, pounds). <Tank ID>
The tank name or location index number.
Example <Level>

A fill tank fills one 10-gallon container every 15 When the tank level rises to this value, ProModel exe-
seconds. After filling, each container moves cutes any subsequent logic.
to a location called FillStation. To model this
activity, define the following activated sub-
routine (this subroutine creates a filled con- Example
tainer every 15 seconds):
Tank_Loop //logic repeats continuously A tanker waits at a dispatch station until the
level of a finished goods tank rises to 2,000
{
gallons. Once the tank level reaches this
WAIT 15 SEC point, a signal dispatches the tanker to the
Tank_Dec(FillTank, 10) finished goods tank for loading. Meanwhile,
the finished goods tank continues filling. To
ORDER 1 Container TO FillStation model this situation, define the following pro-
} cess logic for the tanker at the dispatch sta-
tion:
Tank_RiseTrigger (FGTank, 2000)

See Also
“Emptying to an Entity” on page 211.
Please note

Tank_RiseTrigger Using the Tank_RiseTrigger subroutine instead


of a WAIT UNTIL statement prevents the next
Syntax samples tanker from dispatching until the finished goods
tank falls back below 2,000 gallons.
TANK_RISETRIGGER (<Tank ID>, <Level>)
TANK_RISETRIGGER (TankA, 3000)
See Also
“Defining Trigger Levels” on page 215.
Description
Waits until tank contents rises to a specific level.
Use Tank_RiseTrigger to initiate some action
when a tank rises to a certain level.
200 Chapter 5:
Pre-defined Tank Subroutines

Tank_FallTrigger Please note

Syntax sample Using Tank_FallTrigger instead of a WAIT


UNTIL statement prevents the action from trig-
gering again until the level first rises above the
TANK_FALLTRIGGER (<Tank ID>, <Level>)
fall trigger level.
TANK_FALLTRIGGER (TankB, 500)

See Also
Description
“Defining Trigger Levels” on page 215.
Waits until tank contents falls to a specified level.
Use Tank_FallTrigger to initiate an action when a
tank level falls to a specific level.
Tank_Cap

Components Syntax samples

<Tank ID> TANK_CAP (<Tank ID>)

The tank name or location index number. TANK_CAP (TankA)

<Level>
When the tank level falls to this value, any subsequent
logic executes. Description
Returns the capacity defined for the specified
tank.
Example
Use Tank_Cap when you need to know the
defined capacity for a tank.
When an in-process tank, TankX, falls to 1000
gallons, it triggers a mixing tank to begin pro-
ducing more product. To model this, define Components
the following activated subroutine to control
the mixing tank: <Tank ID>
Tank_Loop //logic repeats continuously The tank name or location index number.
{
Tank_FallTrigger(TankX, 1000)
[Insert logic to mix new batch here] Tank_FreeCap
}
Syntax samples

TANK_FREECAP (<Tank ID>)


ProModel 201
User Guide

TANK_FREECAP (TankA) <Operation time>


The duration (in time units defined in the General
Information dialog) of the operation.

Description
Returns the available capacity of the specified Example
tank.
Use Tank_FreeCap when you need to know the After technicians add all the necessary ingre-
available capacity of a tank. dients to the mixing tank, the tank requires a
20 minute mixing time. To define this opera-
tion, enter the following statement in the sub-
Components routine for the mixing activity:
Tank_DoOperation(MixingTank, 20)
<Tank ID>
The tank name or location index number.

See Also
Tank_DoOperation “Mixing and Reactor Tanks” on page 211.

Syntax samples Tank_GoDown


TANK_DOOPERATION (<Tank ID>, <Operation Syntax samples
time>)
TANK_DOOPERATION (TankA, 30) TANK_GODOWN (<Tank ID>, <Down time>)
TANK_GODOWN (TankA, 5)

Description
Sets the state of the tank to Tank_Operation and Description
waits for the specified operation time. ProModel
Sets the state of the tank to Tank_Down, waits
sets the state to Tank_Blocked after the opera-
for the specified downtime, then sets the state
tion.
back to the previous setting. If you defined a
Use Tank_DoOperation when some activity or downtime using the location downtime dialog,
treatment time is necessary for the material in a call the Tank_GoDown subroutine in the down-
tank. time logic rather than use a WAIT statement. If
the downtime is for cleaning, use the
Components Tank_DoPrep subroutine.
Use Tank_GoDown to shut down a tank due to
<Tank ID> equipment failure (e.g., pump failure). If the
The tank name or location index number. downtime occurs periodically, you can define a
202 Chapter 5:
Pre-defined Tank Subroutines

clock downtime in the downtime logic for the Tank_GoDownSched


tank location and use Tank_GoDown in place of
the WAIT statement.
Syntax samples
Components
TANK_GODOWNSCHED (<Tank ID>, <Down
time>)
<Tank ID>
TANK_GODOWNSCHED (TankA, 5)
The tank name or location index number.
<Down time>
The duration (in time units defined in the General
Information dialog) of the downtime.
Description
Sets the state of the tank to
Tank_ScheduledDown, waits for the specified
Example scheduled downtime, then sets the tank state back
to its previous setting. If you defined a scheduled
A fill line from a dry supply bin plugs randomly downtime using the location downtime dialog,
according to an exponential distribution with a call the Tank_GoDownSched subroutine in the
mean of 10 minutes. The time to unplug the line downtime logic rather than use a WAIT state-
is normally distributed with a mean of 5 minutes ment. If the downtime is for cleaning and you
and a standard deviation of 1 minute. To define will return the tank status to idle, use the
this behavior, define a clock downtime for the bin Tank_DoPrep subroutine.
to occur with a frequency of E(10) minutes. In the
logic defined for the downtime, enter the follow- Use Tank_GoDownSched to shut down a tank for
ing logic: a scheduled task or event (e.g., interim mainte-
nance or end of scheduled workday). Since the
tank uses a scheduled downtime, the time lapsed
during the event does not record as a downtime.
ProModel 203
User Guide

Components Description
Sets the state of the tank to Tank_Setup, waits for
<Tank ID> the specified time, then sets the state to
The tank name or location index number. Tank_Idle. Use Tank_DoPrep for cleaning activi-
<Down time> ties after you empty a tank.

The duration (in time units defined in the General Use Tank_DoPrep to take a tank off line for
Information dialog) of the scheduled downtime. cleaning or other preparation time.

Components
Example
<Tank ID>
Every 4 hours, a technician must check the fill line The tank name or location index number.
from a dry supply bin. The time required to check
the line is normally distributed with a mean of 10 <Prep time>
minutes and a standard deviation of 3 minutes. The duration (in time units defined in the General
To define this behavior, define a clock-based, Information dialog) of preparation time.
scheduled downtime for the bin to occur with a
frequency of 4 hours. In the logic defined for the
downtime, enter the following:
Example

Workers clean a mixing tank for 30 minutes


after each batch produced. To model this,
enter the following logic in the mixing subrou-
tine defined for the mixing tank:
Tank_Loop //logic repeats continuously
{
[Enter mixing and transfer logic here]
Tank_DoPrep(MixingTank, 30)
}

Tank_DoPrep
See Also
Syntax samples “Mixing and Reactor Tanks” on page 211.

TANK_DOPREP (<Tank ID>, <Prep time>)


TANK_DOPREP (TankA, 5)
204 Chapter 5:
Pre-defined Tank Subroutines

Tank_SetState Description
Selects an output tank from among several tanks
Syntax samples based on a selection rule and optional product
type. To use this function, list all tanks included
TANK_SETSTATE (<Tank ID>, <State>) in the selection decision together in the Location
module.
TANK_SETSTATE (TankA, Tank_Idle)

Components

Description <First Tank>


Sets the state of the tank (e.g., Tank_State[<Tank The name or location index number of the starting tank
ID>]) to a new state and updates the statistics in the range.
since the last change of state. <Number of Tanks>
Use Tank_SetState to explicitly change the state The number of tanks in the selection range (limit 10).
of a tank. Use Tank_SetState only if the default <Selection Rule>
state changes do not adequately meet modeling
needs. The rule for making the selection may be one of
the following:
Components Tank_InOrder (selects the first idle tank
encountered)
<Tank ID> Tank_LongestIdle (selects the tank idle
the longest)
The tank name or index number.
<Maximum Level>
<State>
The maximum level of the output tank before consider-
The new state for the tank. For a list of possible tank
ing it for selection. Enter 0 if the output tank must be
states, see “Tank States” on page 190.
empty or idle before being considered.
<Product Type>
An integer specifying the required value of the Product
Tank_SelectOutput array in order to select the tank. Enter 0 if the tank
selection requires no product type match. (This applies
Syntax samples only if the maximum level specified is greater than 0.)

TANK_SELECTOUTPUT (<First Tank>, <Number


of Tanks>, <Selection Rule>, <Maximum Example
Level>, <Product Type>)
TANK_SELECTOUTPUT (TankA, 3, Tank_InOrder, A supply tank feeds one of 3 output tanks
5000, 0) and always gives preference first to Tank1,
then to Tank2, and finally to Tank3 based on
availability. Furthermore, the supply tank can
select a tank only if its contents are less than
ProModel 205
User Guide

8000 gallons. To model this selection, list .


Tank1, Tank2, and Tank3 together (and in
SelectedTank=Tank_SelectOutput(Tank1, 3,
order) in the location module. Then define
Tank_InOrder, 8000,0)
the following statement to select the tank
using a local variable, Selected_Tank:
INT Selected_Tank
. The diagram below shows the logic used to make
a tank selection:
.

No
Is the specified Max
level > 0?

Yes

Is
there a tank Is
No No Wait for first tank to
No there an idle
with level > 0 but Is the specified Max become idle.
less than Max level and tank that satisfies
level > 0?
same product selection rule?
type?

Yes Yes
Yes

Match Wait for first tank to


fall below the Max
specified product
type if required level.
No

Yes

Return Tank ID

If you base a tank selection on product type, you See Also


must be careful to assign an appropriate integer
“Selecting from Multiple Input or Output Tanks”
value to the Product array element corresponding
on page 213.
to the tank location.
206 Chapter 5:
Pre-defined Tank Subroutines

Tank_SelectInput An integer specifying the required value for the Prod-


uct array in order to select the tank. Enter 0 if the tank
selection requires no product type match.
Syntax samples

TANK_SELECTINPUT (<First Tank>, <Number of Example


Tanks>, <Selection Rule>, <Minimum Level>,
<Product Type>)
A tanker arrives at a pick up station to load
TANK_SELECTINPUT (Tank1, 5, Tank_ByOrder, from one of 5 tanks depending on which tank
1000, 0) has been full the longest. The tanker will fill
from a partial tank if the tank has any con-
tents at all (at least .1 gallons). To model the
tank selection, define the following operation
Description logic for Tanker at PickUpStation:

Selects an input tank from among several tanks INT SelectedTank


based on a selection rule. To use this function, Selected_Tank=Tank_SelectInput(Tank1, 5,
you must list all tanks included in the selection Tank_LongestBlocked, .1, 0)
together in the Location module.

Components
See Also
<First Tank> “Selecting from Multiple Input or Output Tanks”
on page 213.
The name or location index number of the starting tank
in the range.
<Number of Tanks> Tank_UpdateStats
The number of tanks in the selection range (limit 10).
<Selection Rule> Syntax samples
The rule for making the selection may be one of the
TANK_UPDATESTATS (<Tank ID>)
following:
TANK_UPDATESTATS (TankA)
Tank_InOrder (selects the first blocked
tank encountered)
Tank_LongestBlocked (selects the tank
blocked the longest) Description
<Minimum Level>
ProModel calls this subroutine automatically
The minimum level of the input tank before consider- whenever you call any of the predefined subrou-
ing it for selection. If the tank must be full before con- tines that affect the tank level. If you change the
sidering it for an input source, enter 0. value of the Tank_Level directly, call the
<Product Type> Tank_UpdateStats subroutine afterward. This
subroutine updates the current statistics on the
tank and sets the state to Tank_Filling (if filling),
ProModel 207
User Guide

Tank_Emptying (if emptying), Tank_Blocked (if The name or location index number of the TO tank
full), or Tank_Idle (if empty). (this value should be 0 if there is no TO tank).

You should not need to use this subroutine unless


you defined a customized Tank_Empty,
Tank_Fill, or Tank_Transfer subroutine.
Example

TankA fills with 10,000 gallons at a rate of 60


Components gpm until it reaches a level of 9,700 gallons.
Then it fills at a rate of 30 gpm. To model this
<Tank ID> change of rate, define the following logic in
the Tank_Rate subroutine:
The tank name or location index number.
IF Tank_ToID=TankA
THEN IF Tank_Level[TankA]<9700
Tank_Rate THEN RETURN 60
ELSE RETURN 30
Syntax sample Now when you fill TankA, enter the following:
Tank_Fill(TankA, 10000, 0, 0)
TANK_RATE (<FROM tank ID>, <TO tank ID>)
The first 0 in the expression above causes the
TANK_RATE (TankA, TankB) logic defined in the Tank_Rate subroutine to
execute and determine the flow rate.

Description
ProModel calls this subroutine automatically if See Also
you pass a 0 value as the From Rate when using “Varying the Transfer Rate” on page 214.
the Tank_Empty or Tank_Transfer subroutine.
To return the desired rate value, enter the neces-
sary logic in the subroutine—ProModel calls the Pre-defined Data Elements
subroutine with each time step. A return value of
The ProModel tank submodel provides the fol-
0 terminates the flow and returns the remaining
lowing data elements for modeling tanks. Unless
amount in the Tank_QtyLeft variable.
otherwise specified, all arrays are single-dimen-
sional and of type integer. Initially, these arrays
Components are 100 elements in size to allow for up to 100
locations. If you define more than 100 locations,
<FROM Tank> you will need to enlarge the array or place tanks
The name or location index number of the FROM tank toward the beginning of the location list (within
(this value should be 0 if there is no FROM tank). the first 100 locations).
<TO Tank> Tank_Level array Stores the level of each tank.
Since the values in this array directly control the
208 Chapter 5:
Pre-defined Tank Subroutines

tank gauge and tank statistics, the array MUST be Location States by Percentage. When reporting
present in every tank model. Location statistics for tanks, note the following:
Tank_State array Tracks the state of the tank. •Total Entries The number of units (e.g.,
gallons, pounds) to enter the tank.
Tank_Product array An optional array used to
•Avg Minutes Per Entry Left blank since
record or test the product currently at a tank.
there is no individual entry for a tank.
Tank_Statistics array A two-dimensional array
Tank_Fills array An optional array used to track
of type real used to record tank level statistics
the number of transfers to a tank. This is espe-
whenever the level changes. Generally, you will
cially useful when you activate multiple
never need to reference this array since values
Tank_Fill or Tank_Transfer subroutines for a
automatically update when you use the pre-
tank and you wish to know when the fills are
defined Tank subroutines. All times are in time
complete. The user sets the value of Tank_Fills to
units defined in the General Information dialog.
zero before activating the subroutines, then
ProModel always gathers these statistics but
defines a WAIT UNTIL statement after the
reports them only if you check Basic or Time
ACTIVATE statement. The Tank_Fills array
Series statistics for the tank location.
increments automatically when a Tank_Fill or a
Column Description Reset After Warm-up Tank_Transfer subroutine executes. See “Mix-
1 Last level NC ing and Reactor Tanks” on page 211 for addi-
2 Last change time Current time in time tional information.
units defined in the
General Information
dialog Statistics
3 Cum time-weighted 0
level
4 Entries Value of column 1
Please note
5 Max contents Value of column 1
6 Last State Change Current time in time
TS = Tank_Statistics array
units defined in the n = Location index number of tank
General Information
dialog
7 Cum time Idle 0
8 Cum time Operation 0
Calculating Location Statistics for
9 Cum time Setup 0
Tanks
10 Cum time Filling 0
11 Cum time Emptying 0
Entries = TS [n, 4]
12 Cum time Blocked 0
13 Cum time Down 0 Avg. Time per Entry = (not applicable)
14 Current downtime NC Avg. Contents = TS[n,3] / Scheduled Time
count
Max Contents = TS [n,5]
As shown in the previous table, the statistics col- Current Contents = TS [n,1]
lected in the Tank_Statistics array automatically Utilization=100 x TS[n,3] / (Capacity x Sched-
reset after any warm-up period. ProModel reports uled Time)
output statistics under Location statistics and
ProModel 209
User Guide

Calculating Location State Statis- be capable of executing independently of each


tics for Tanks other. On the other hand, if the logic associated
with two tanks is interdependent, only one con-
trol subroutine is necessary. For example, if
%Operation = 100 x TS[n, 8] / Scheduled Time
TankA fills TankB while TankB waits, then
%Setup = 100 x TS[n, 9] / Scheduled Time TankB pumps out while tank A waits, you need
%Idle = 100 x TS[n, 7] / Scheduled Time only a single control subroutine since you control
both tanks by a single logic sequence. If a single
%Waiting = 100 - Sum of other percentages
tank feeds several other tanks independently, you
%Blocked = 100 x TS[n, 12] / Scheduled Time would need a separate subroutine to control each
%Down = 100 x TS[n, 13] / Scheduled Time output. In most cases, you will need at least one
control subroutine per tank and, in certain situa-
tions, you may wish to use a hierarchical control
system (i.e., a master or supervisory control sub-
To gather statistics on how much of a particular routine) to activate subordinate subroutines.
product was processed, you may define variables
to record product processing during the simula- Most tank control subroutines should be activated
tion. subroutines. In contrast to called subroutines,
activated subroutines use the ACTIVATE state-
ment and cause the logic activating the subrou-
Defining Tank Control Subrou- tine to continue independently of the activated
tines subroutine. This allows you to execute multiple
control subroutines concurrently. Multiple tanks
Unlike defining entity activity at a location with identical control logic may share the same
(defined in the Processing module), modeling control subroutine if you activate the subroutine
tank location activity requires the use of subrou- for each tank and pass the tank ID as a parameter.
tines. Many of these subroutines are user-defined
One of the keys to modeling interactive tank
and called using the ACTIVATE statement.
behavior is to effectively use WAIT UNTIL
Though you generally activate them from the ini-
statements. When you use WAIT UNTIL state-
tialization logic, you may also activate them from
ments based on the Tank_Level array, use them
another tank subroutine. Tank subroutines consist
sparingly since this array changes frequently and
of logic defined to control when, where, and how
may slow the simulation.
much to empty, fill, or transfer from a tank.
Often, these subroutines require the use of WAIT
UNTIL statements to monitor conditions (e.g., Examples of Tank Control Logic
the tank level or state) before making a transfer
and may include delays for mixing or cleaning. The following examples show how to model dif-
ferent tank and flow situations. For full models
At a minimum, you should define a separate con- illustrating these situations, see the reference
trol subroutine for any logic that executes inde- model in the MODELS\REFS directory within
pendently of any other logic. For example, if the ProModel directory.
TankA fills TankB at the same time TankB trans-
fers to some other tank, you should define two
separate subroutines since both sets of logic must
210 Chapter 5:
Pre-defined Tank Subroutines

Filling from an Entity attribute adds to the ReceivingTank. Once the


tanker delivers this quantity, it is free to execute
A typical tank modeling situation is the arrival of the routing defined for it at the Delivery location.
an entity (e.g., a tanker or other vehicle) to
deliver its contents to a tank. To model this situa-
tion, define an arrival or routing for the entity, Initializing and Replenishing
causing it to enter the location where it will make Supply Tanks
its delivery. In the entity processing logic at the A supply tank is an originating tank that is a
delivery location, call the Tank_Fill subroutine. source of raw material for one or more down-
By calling rather than activating the subroutine, stream tanks. Often, supply tanks contain ingredi-
you will detain the delivering entity until ents that feed into a mixing tank or hold
Tank_Fill executes. Note that the material does chemicals that feed into a reactor. Typically, you
NOT route from the delivery location to the tank. replenish a supply tank when it gets low and
Instead, the Tank_Fill subroutine simply fills the make it available for use whenever it has an ade-
tank with a specified quantity while the entity quate supply. If you always stock the supply tank
waits. Unless the quantity is a constant amount, it and it is always available for use, you do NOT
is usually a good idea to use an entity attribute to need to model it since it poses no constraint on
store this quantity value. After filling the contents the process. You may set supply tanks to an ini-
into the tank, the entity is free to continue pro- tial level at the start of the simulation in the ini-
cessing. tialization logic, then use them as needed by a
To illustrate how an entity might transfer its con- mixing or other downstream tank. To initialize
tents to a tank, suppose an entity, Tanker, arrives the level in a supply tank, enter the following
at a location, Delivery, carrying a quantity of gal- statement in the initialization logic:
lons stored in an entity attribute called
Tanker_Qty. The tanker discharges its contents Initialize tank level
into a tank, ReceivingTank, at a rate of 200 gal-
lons per minute. Once the ReceivingTank Tank_SetLevel (<supply tank>, <qty>)
becomes full, the level must drop to 1000 gallons
before filling resumes. Since the entity is tied up
while it discharges into the tank, use the follow-
ing statement in the processing logic for Tanker If, for example, you wanted to begin the simula-
at Delivery to define the logic used to fill the tion with the supply tank, WaterTank, filled with
tank: 800 gallons of water, you would enter:

Fill the tank Start with full tank

Tank_Fill (ReceivingTank, Tanker_Qty, 200, Tank_SetLevel (WaterTank, 800)


1000)

To gradually fill or refill a supply tank whenever


The above statement causes each arriving tanker it drops below a trigger level, use the Tank_Fill
to wait until the quantity stored in its Tanker_Qty subroutine with a large fill quantity and an appro-
ProModel 211
User Guide

priate resume level. For example, the following Tank_Transfer (Tank2,MixingTank, 300, 20, 4, 0)
statement will continue pumping up to 999999 Tank_DoOperation (MixingTank,15) //Mix time
units into TankA at a rate of 200 units per minute.
Wait Until Tank_State [StorageTank]=
Whenever the tank becomes full, it must drop to
Tank_Idle /* Waits for storage tank availability
400 units before filling resumes.
*/
Tank_Transfer (MixingTank, StorageTank,
Resume fill with trigger level Tank_Level[MixingTank],40, 0, 0)
Tank_Prep (MixingTank, 50) // Clean mixing
Tank_Fill (TankA, 999999, 200, 400)
tank for 50 minutes.
END

Mixing and Reactor Tanks


Mixing and reactor tanks receive material usually If the ingredients feed into the mixing tank at the
from one or more supply tanks. Once it receives same time rather than sequentially, activate the
the material, the tank may require a mixing or Tank_Transfer subroutines for the mixing tank
other reaction time. To illustrate, suppose we and monitor the Tank_Fills array to know which
have two tanks (Tank1 and Tank2) supplying ingredients enter into the tank. For simultaneous
ingredients to a tank called MixingTank. First, fills, replace the first two transfer statements fol-
workers pump 2000 gallons of a liquid from lowing the BEGIN statement in the previous sub-
Tank1 at 50 gallons per minute followed by the routine with the following logic:
transfer of 300 pounds of dry mix from Tank2 at
20 pounds a minute (the dry mix adds .2 gallons
to the level of the MixingTank for every pound
Simultaneously mix, then clean
transferred, equating to 4 gallons per minute). tank
The ingredients then mix for 15 minutes before
transferring to an idle storage tank. After trans- Tank_Fills[MixingTank]=0
ferring the mix, workers must clean the Mixing- ACTIVATE Tank_Transfer(Tank1, MixingTank,
Tank for 50 minutes to prepare it for the next 2000, 50, 0, 0)
mixing cycle.
ACTIVATE Tank_Transfer(Tank2, MixingTank,
The control logic for the mixing tank should be a 300, 20, 4, 0)
subroutine activated from the initialization logic WAIT UNTIL Tank_Fills[MixingTank]=2
which continues to loop throughout the simula-
...
tion. The subroutine logic might appear as fol-
lows:

Mix and clean the tank Emptying to an Entity


Often, tanks deliver material to discrete entities
Tank_Loop //logic repeats continuously
such as containers (or perhaps the material itself
BEGIN converts to discrete entities through a solidifica-
Tank_Transfer (Tank1,MixingTank,2000, 50, 0, 0) tion or consolidation process). In either case, you
212 Chapter 5:
Pre-defined Tank Subroutines

can draw from the delivery tank using the Tank Transfers
Tank_Empty subroutine if outflow is gradual and
defined by a flow rate, or the Tank_Dec subrou-
tine if the output occurs in discrete intervals
based on a bottling or packaging time.
To output material from a tank without modeling
the entity to which it outputs, call the
Tank_Empty or Tank_Dec subroutine. To trans-
fer material from a tank to entities arriving at a
filling station (remember, the filling station itself
is NOT a tank), route the entities to the filling sta-
tion using a SEND or other routing rule, then call
the Tank_Empty or Tank_Dec subroutine.
If using the Tank_Dec subroutine, the entity When transferring from one tank to another, you
should wait for the fill time before decreasing the must determine whether the source tank makes
tank level since Tank_Dec happens instantly. For the decision to transfer to the destination tank (a
example, if a bottling operation fills a 2 gallon push approach) or whether the destination tank
container every 6 seconds, define the following makes the decision to draw material from a
processing logic for the container at the fill sta- source tank (a pull approach). You should define
tion: a control subroutine from the perspective of the
tank that makes the decision. If the model
Wait to fill from tank, then con- requires no tank selection, specify a WAIT
tinue UNTIL statement to wait until the FROM or TO
tank satisfies the condition required for transfer.
For example, if a source tank makes the decision
Wait 6 sec
to transfer to a destination tank whenever the des-
Tank_Dec (Filler, 2) tination tank becomes idle, enter the following
statement in the subroutine:

If the delivery tank has insufficient contents to Transfer contents when idle
decrease the level by the specified amount, the
processing will automatically pause until enough Wait Until Tank_State [<destination tank ID>] =
material is available. Once the specified quantity Tank_Idle
empties, the entity can continue processing. To
create an entity as the result of an emptying oper-
ation, define an activated subroutine that empties
If the destination tank makes the decision to
the desired quantity, then execute an ORDER
transfer (a pull approach), you should base the
statement. This will create a new entity at the fill-
WAIT UNTIL statement on a required condition
ing station.
for the source tank as follows:
ProModel 213
User Guide

Transfer contents based on condi- Selecting from Multiple Input or


tion Output Tanks
To enable one or more tanks to select from sev-
Wait Until Tank_State [<source tank ID>] = eral input or output tanks, use the pre-defined
Tank_Blocked
subroutine Tank_SelectInput or
Tank_SelectOutput (see subroutine descriptions).
ProModel bases tank selection on which tank is
Following the WAIT UNTIL statement, call the ready to transfer or receive and that has the same
Tank_Transfer, Tank_TransferUpTo, or ProductType array value. You must list the tank
Tank_TransferDownTo subroutine to transfer selections together in the Location module.
from the source tank to the destination tank. For example, if TankX selects from among three
To illustrate how to define a tank transfer using a input tanks (Tank1, Tank2, and Tank3) based on
pull approach, suppose that TankB requires 1000 which input tank has waited the longest to dis-
gallons from TankA whenever TankB becomes charge its contents, you would enter the follow-
empty. TankB will draw material from TankA ing logic in the control subroutine defined for
only when TankA has a minimum level of 1000 TankX:
gallons. The subroutine to define this logic might
appear as follows: Select from multiple tanks

Tank transfer in a pull system Int SelectedTank


SelectedTank = Tank_SelectInput(TankX,
Tank_Loop //logic repeats continuously Tank1, 3,
Tank_LongestBlocked, 0)
Begin
Wait Until Tank_Level[TankA] >= 1000 /*Wait
for TankA to reach 1000 gallons*/
Tank_Transfer (TankA, TankB, 1000, 200,0,0) /* The first statement defines a local variable,
Transfer 1000 gal to TankB at 200 gpm*/ SelectedTank, used to assign which tank you
select. The second statement calls the SelectInput
[Enter TankB processing and emptying logic
subroutine specifying that TankX is to select one
here]
of three tanks beginning with Tank1.
End Tank_LongestBlocked causes TankX to select
the tank blocked the longest (i.e., tank is full or
waiting). Entering 0 at the end prevents selecting
a full tank. If no tank is full, the statement does
Please note not execute until one of the input tanks fills. With
a tank ID assigned to SelectedTank, you can call
To select from among multiple input or output a transfer subroutine to make the transfer.
tanks, activate this subroutine in the initialization
logic.
214 Chapter 5:
Pre-defined Tank Subroutines

For output tanks, you would define similar logic The subroutines TransferToB and TransferToC
but include Tank_SelectOutput instead of would each execute a Tank_TransferDownTo
Tank_SelectInput. command followed by a statement incrementing
the value of TransferDone. For example, the
logic for TransferToB would be as follows:

Split transfers subroutines

TransferDownTo(TankA, TankB, 0, 30, 0)


INC TransferDone

Varying the Transfer Rate


The transfer or empty rate can change dynami-
Split Transfers cally during an empty, fill, or transfer. To vary
the rate of flow, pass 0 as the flow rate when call-
Sometimes it is necessary to use a tank or separa- ing any of the transfer, fill, or empty subroutines.
tor to split the flow to several output tanks. To This calls the Tank_Rate subroutine automati-
define the concurrent transfer of material from cally with each time step. You should modify the
one tank to multiple tanks, define an activated Tank_Rate subroutine so that it returns the appro-
subroutine for each transfer. Suppose, for exam- priate rate value.
ple, that when TankA fills it begins transferring
to TankB at a rate of 30 gpm and to TankC at a Suppose, for example, that TankA transfers to
rate of 40 gpm. To know when both transfers are TankB at a rate that decreases from 150 gpm to
complete, define a global variable (e.g., Transfer- 50 gpm when the level of TankB reaches 4000.
Done) which increments at the end of each trans- To achieve this, pass 0 as the From Rate when
fer. Defining the following logic would initiate you call the transfer subroutine, then enter the
this split transfer once TankA is full: following logic in the Tank_Rate subroutine:

Initiate split transfer Vary the transfer rate

ACTIVATE TransferToB () // initiates transfers IF (Tank_FromID = TankA) AND (Tank_ToID =


from A to B TankB)

ACTIVATE TransferToC () // initiates transfers THEN IF Tank_Level[TankB] >= 4000


from A to C THEN RETURN 50
WAIT UNTIL TransferDone = 2 // Wait until trans- ELSE RETURN 150
fers are complete
TransferDone = 0 // reset for next transfer
ProModel 215
User Guide

Dynamically Suspending Flow associated action should execute when the tank
level rises or falls to a certain level.
To momentarily interrupt flow into or out of a
tank, use the Tank_GoDown subroutine or set the
state of the tank to down (Tank_SetState =
Tank_Down). This typically happens if a pump
fails but may occur in other situations.

Dynamically Terminating a Flow


Normally, the flow into or out of a tank stops
once you reach the desired quantity or level.
However, in some situations you may wish to ter-
minate a transfer if some event or condition
occurs that you cannot predetermine (e.g., a deci-
sion to divert flow to a preferred outlet tank that To show how to define a trigger subroutine, sup-
just became available). In this case, you can turn pose that whenever TankA rises to 2000 gallons,
off the flow into or out of a tank by specifying a an entity called Truck travels to a location called
variable transfer rate instead of a fixed transfer Pickup. The logic for this trigger subroutine
rate (see previous discussion, Varying the Trans- might look as follows:
fer Rate). A variable transfer uses the Tank_Rate
subroutine to determine the rate for each time Trigger subroutine
step—to terminate a transfer, return a rate value
of 0. Tank_Loop //logic repeats continuously
Begin
Defining Trigger Levels
Tank_RiseTrigger (TankA, 2000) /* waits for
A trigger level is a level to which material in a TankA to rise to 2000 units*/
tank either falls or rises and triggers some action. Order 1 Truck to Pickup // order a Truck to
To continuously monitor when a tank reaches a Pickup
trigger level, define and activate a trigger subrou-
End
tine in the initialization logic. The subroutine
should call Tank_RiseTrigger or
Tank_FallTrigger depending on whether the
Once the tank reaches the trigger level, the
Tank_RiseTrigger subroutine prevents further
triggering until the level drops back below the
trigger level first.
When you use trigger subroutines, use them spar-
ingly because they are CPU intensive. Every time
the tank level changes, ProModel tests to see if
the tank reached the trigger level. Trigger subrou-
tines are often unnecessary because, unlike an
216 Chapter 5:
Pre-defined Tank Subroutines

actual tank where sensors report the tank level, Showing Pipes
you directly control how much to pump into a
tank. For instance, an alternative way to model To show pipes connecting the tanks, use paths or
the previous example without using a triggering background graphics. If you desire to show the
subroutine would be to call the material in the pipe, use a long, skinny tank with
Tank_TransferToLevel subroutine to first fill the a capacity of 1 to represent the pipe. You can set
tank to the 2000 unit level, order the Truck entity the level of this tank to 0 or 1 to show product
and then transfer the rest. flow. For example, suppose we define a tank
location called Pipe used to represent the connec-
tion between Tank1 and Tank2. Whenever trans-
Processing Multiple Products ferring from Tank1 to Tank2, you would enter
Where you must track several different products the following:
through one or more tanks, it may be useful to
define macros for naming each product type. For Define pipes
example, setting ProductA equal to 1 and Pro-
ductB equal to 2 will improve the readability of Tank_SetLevel (Pipe,1)
the model. To track which product a particular
Tank_Transfer (Tank1, Tank2, ….)
tank is processing, ProModel uses a pre-defined
integer array called Tank_Product—the user is Tank_SetLevel (Pipe, 0)
responsible for maintaining the array values. If,
for example, ProductA begins pumping into
Tank1, enter the following after you assign an
integer value to ProductA in the Macros module High-Rate Entity Processing
to distinguish it from other products: For systems that process entities at rates higher
than one hundred units per minute, using discrete
Tracking products entities could make the simulation extremely
slow. For this reason, ProModel uses tanks. To
Tank_Product [Tank1] = ProductA use a tank to model high-rate processing, think of
the tank as a buffer where the tank level repre-
sents the number of items in the buffer. For
example, suppose that bottles feed through a fill-
ing station at a rate of 110 per minute. The input
buffer, FillerInput, has a capacity of 1200 bottles
and the output buffer, FillerOutput, has a capacity
of 2000 bottles. If FillerOutput is full, processing
stops until the quantity in the output buffer drops
to 1500 bottles. An arriving container feeds quan-
tities of 200 bottles to the FillerInput location and
it takes 1 minute to unload the container. When
the filling station fills 50 bottles, workers put the
bottles into a box (represented by an entity) and
ship them. Since workers load the boxes as soon
ProModel 217
User Guide

as the bottles complete the filling process, there is Order 1 Box to Shipping
no delay time involved. }
The operation logic for the container at the arriv-
ing location would be as follows:
Please note
High-rate processing
The Tank_Dec statement automatically removes
WAIT 1 min
50 bottles from FillerOutput whenever there are
Tank_Inc (FillerInput, 200) at least 50 bottles available.

To model the processing of bottles from FillerIn- Special Notes


put to FillerOutput, enter the following statement
in the model initialization logic. •Since tank models do not stop automatically
when there are no more entities or sched-
uled arrivals, remember to define a run
High-rate processing length or a STOP statement.
•When you activate a subroutine, it doesn't
ACTIVATE Tank_Transfer(FillerInput, FillerOut- process until the current logic (the one
put, 999999, 110, 0, 1500)
activating the subroutine) finishes or
becomes blocked. It you want the acti-
vated subroutine to process first, enter
This statement causes the FillerInput tank to “WAIT 0” after the ACTIVATE state-
transfer bottles to FillerOutput at a rate of 110 per ment.
minute whenever there are bottles in FillerInput •Do not define a local variable inside of a
and capacity available in FillerOutput. The Tank_Loop since the loop will create the
resume level is 1500. (Up to 999999 bottles will variable multiple times.
transfer.) •Make sure all IF...THEN logic and WAIT
UNTIL statements based on the
To model the creation of a 50-bottle box each Tank_Level array use the “>=” or “<=”
time the filling station fills 50 bottles, define and operator and not just an “=” operator.
activate the following subroutine in the model (This is because flow occurs in increments
initialization logic: and you can’t check for an exact value.)
•Tanks are not legal in multi-unit locations or
Create new, combined unit in locations containing a conveyor or
queue.
Tank Loop //causes logic to repeat continu-
ously
{
Tank_Dec(FillerOutput, 50)
218 Chapter 5:
Background Graphics

Background Graphics How to create or edit background


graphics:
Background graphics allow you to enhance a
model by adding a background to the animation.
1. Select Background Graphics from the
A background could show a floor-plan of a fac- Build menu.
tory or any item that is not part of a location,
entity, or resource. Backgrounds can be created 2. Select Front of Grid or Behind Grid
depending on the mode desired.
using the tools in the Background Graphics Edi-
tor or by importing an existing background from
another application, such as AutoCAD. Imported
graphics must have been saved in one of the fol-
lowing formats BMP, WMF, GIF, or PCX. The
Background Graphics Editor
Background Graphics Editor is accessed from the Modes
Build menu as shown below.
ProModel gives you the option of placing the
background graphic in front of or behind the grid.
This is useful when you want to view the
imported background graphic, but you also want
to see the grid for drawing and sizing objects on
the imported background graphic.

Front of Grid Mode


Creating graphics in this mode places the back-
ground graphic in front of the grid as shown
below.

There are two modes for editing in the Back-


ground Graphics Editor: Front of Grid and
Behind Grid. Generally, most graphics should be
laid out in front of the grid. Graphics placed
behind the grid should be reserved for large
objects such as walls or imported backgrounds.

Behind Grid Mode


Creating graphics in this mode places the back-
ground graphic behind the grid as shown below.
ProModel 219
User Guide

located in the lower right portion of the


workspace.

Library Graphics Window


The Library Graphics window contains the icons
of the current graphic library file, specified in the
General Information dialog. These icons may be
placed on the Layout in the same way as other
objects.

Background Graphics Editor


The Background Graphics Editor allows you to You may size the window as desired, or use the
place icons, text and other graphic shapes on the scroll bar shown above to scroll through the
layout behind locations and other system element available icons.
graphics. The arrangement of the two windows
and button bar is shown below. Tools Button Bar
The Tools button bar contains the tools necessary
to create objects various shapes. It also contains

•The Library Graphics window, containing


all the icons in the current graphic library,
is located at the top of the workspace.
•The Tools button bar, where you may select
a tool for creating and editing graphic
shapes, is located at the left of the work-
space.
•The Layout window, where all creating and
editing of graphic shapes is done, is
220 Chapter 5:
Background Graphics

tools for editing those objects including flip, The first four functions apply to the currently
rotate, and cut. selected object. To select multiple objects, hold
the shift key while selecting an object. Alterna-
Selector tively you can drag a rectangle encompassing the
objects you want selected. To deselect one of
several selected objects, click on the selected
object while holding the shift key.

Drawing Tools

Cut Removes the selected object(s) and makes a


temporary copy that may be pasted back into the
Editing Tools
edit window later.
Copy Makes a temporary copy of the selected
object(s) for later pasting.
Paste Adds the most recently cut or copied
object(s) to the Layout window.
The Background Graphics Editor Tools button Delete Deletes the selected background graphic
bar is the nearly the same Tools button bar used from the Layout window.
in the Graphic Editor. The only difference is that
Select All Selects all of the graphic objects in
the Background Graphics Editor Tools button bar
the current mode.
does not contain an entity spot or status light tool.
For more information on the Tools button bar, Copy to Clipboard Copies all graphic objects
see “Graphic Tools Button Bar” on page 323. in the current mode to the clipboard as a bitmap
or windows metafile so they can be pasted into
Edit Menu another application such as a word processor.

Use the Edit menu for selecting and duplicating Paste WMF Pastes a Windows metafile (WMF)
the graphic objects in the current Background from the Windows clipboard into the Layout win-
Graphics mode. You may also use it to exchange dow. You must have previously copied a Win-
graphics with other applications. To use the Edit dows metafile to the Windows clipboard.
menu functions, select the object you wish to edit Paste BMP Pastes a bitmap file (BMP) from the
by clicking on it in the Layout window. Windows clipboard into the Layout window. You
ProModel 221
User Guide

must have previously copied a bitmap file to the How to move an imported back-
Windows clipboard. ground graphic:
Import Graphic Imports a WMF, BMP, PCX,
or GIF file into the Layout window. • Place the cursor on the imported back-
ground graphic and drag it to the desired
Export Graphic Exports all graphic objects in location in the layout.
the current mode to a WMF or BMP file.

Importing a Graphic
How to size an imported back-
Importing a background graphic can bring reality
into the model. For example, if a layout is created
ground graphic:
in a graphic package, it may be desirable to
import the entire layout rather than create it in 1. Place the cursor on one of the four corners
of the imported background graphic. The
ProModel. This is done by saving the file in a
cursor becomes a cross-hair at this point.
graphic package, such as AutoCAD, as a WMF,
BMP, PCX, or GIF file and importing the graphic 2. Drag the cursor to size the background
into ProModel. This can save you an extensive graphic as desired.
amount of time.

How to import a background Please note


graphic into the layout:
Once imported, the background graphic is not a
1. In a graphics application, save the separate file from the model. It is included in the
graphic in one of the following formats, WMF, model. Therefore, when moving or copying a
BMP, PCX, or GIF.
model file from one directory to another, it is not
2. Select Background Graphics from the necessary to move or copy the imported back-
Build menu. ground graphic file as well. On the other hand, if
3. Select Front of Grid or Behind Grid. the external graphic file is changed, it must be re-
imported to update the model layout.
4. Select Import Graphic from the Edit Menu.
5. Enter the name of the graphic you would
like to import.
6. Select OK to close the import graphic dia- Exporting a Graphic
log box. The graphic will then appear in the In some cases, it is desirable to export a graphic
layout window. The upper left corner of the
created in ProModel for use in another applica-
imported graphic will align with the upper left
tion. ProModel will export all objects in the cur-
corner of the layout window.
rent mode (In Front of Grid and Behind Grid) as
one graphic to a WMF or BMP file.
222 Chapter 5:
Background Graphics

How to export a graphic: Rotate Rotates the selected object(s) 90 degrees


clockwise. This does not apply for non-true-type
1. Select Export Graphic from the Edit menu. fonts.
2. Enter a valid DOS name for the graphic in Behind Grid Moves the selected object in the
the resulting dialog box, such as forklift.bmp. layout window behind the grid. Once this is done,
3. Click the OK button in the Export Graphic you must go to Behind Grid mode to edit the
dialog box. graphic.
Front of Grid Moves the selected object in the
layout window in front of the grid. Once this is
done, you must go to Front of Grid mode to edit
Graphics Menu the graphic.
The Graphics Menu allows you to flip and rotate Group Combines or groups several graphic
the selected graphic object(s) in the layout win- objects into a single graphic so they may be sized
dow. It also allows you to specify whether you and edited together.
want the graphic to be behind or in front of the
grid. Additionally, it allows you to group selected Ungroup Ungroups several grouped graphic
graphic objects together into a single graphic. objects so they may be edited individually.
Finally, it provides the option to define line
Lock Locks a graphic in place so that it can’t be
styles, fill patterns, line color, and fill color.
moved. This is helpful for preventing accidental
moving of a graphic that you wish to leave sta-
tionary.
Unlock Unlocks a locked graphic, allowing it
to be moved on the layout.
Alignment When multiple objects are selected
on the layout, they can be aligned side to side, top
to top, etc.
Line Styles Allows the user to define the line
style including transparent, dashed, line thick-
ness, and optional arrowheads on either end of
the line. If any objects are selected, the line styles
of the selected objects are changed.
Fill Patterns Allows the user to define the fill
pattern for solid objects including slant, grid,
Flip Horizontal Horizontally flips the selected crosshatch, backward slant, horizontal, vertical,
object(s). transparent, solid, vertical gradient, and horizon-
Flip Vertical Vertically flips the selected tal gradient. If any objects are selected, the fill
object(s). patterns of the selected objects are changed.
Line Color Allows the user to define the line
color and create custom colors. If any objects are
ProModel 223
User Guide

selected, the line color of the selected objects are


changed.
Fill Color Allows the user to define the fill color
and create custom colors for solid objects. If any
objects are selected, the fill color of the selected
objects are changed.

Please note

All functions in the Graphic menu of Background


Graphics are nearly the same functions described
in the Graphic Editor Graphic menu. Differences
are noted below for moving a graphic behind the
grid and in front of the grid. See “Graphic Edi-
tor” on page 312 for more information on the
functions above.

How to move a graphic behind


the grid:

1. Select the graphic on the layout using the


selector.
2. Select Behind Grid from the Graphics
menu.

How to move a graphic in front of


the grid:

1. Select the graphic on the layout using the


selector.
2. Select Front of Grid from the Graphics
menu.
224 Chapter 5:
Background Graphics
ProModel 225
User Guide

Chapter 6: Building the


Model: Advanced Elements

Attributes How to create and edit attributes:


Attributes are place holders similar to variables, 1. Select Attributes from the Build Menu.
but are attached to specific locations and entities
and usually contain information about that loca-
tion or entity. Attributes may contain integers or
real numbers. You may also assign model ele- Attribute Types
ment names (e.g., StationA) to an attribute, which
Attributes are classified as follows:
is stored as the element’s index number but may
be referenced by name. Attributes are defined
through the Build menu as shown below. Entity Attributes
Entity attributes are place holders assigned to an
entity and contain numerical information about
that entity. An entity attribute is identified by its
name and may be assigned a value or model ele-
ment name stored as a value. An entity attribute
may be examined and acted upon in any of the
following places:
•Arrival logic
•Operation logic
•Move logic which refers to the attribute of
the entity being routed
•Min or Max attribute rules for locations and
resources
•Routing quantity
•Routing destination priority
•Resource pick up and drop off times
•Entity speed
•Debug condition
226 Chapter 6:
Attributes

Location Attributes would be more appropriate to use a local vari-


able. Local variables act like temporary attributes
Location attributes are place holders assigned and are valid only within the logic in which they
directly to a location and contain numerical infor- are defined (see “Variables” on page 231).
mation about that location. A location attribute is
identified by its name and may be assigned a
value or model element name stored as a value. Cloning Attributes
A location attribute may be examined and acted
on in any of the following places: Whenever one entity initiates the creation of
other entities through a SPLIT AS, CREATE, or
•Arrival logic ORDER statement, or as the result of specifying
•Operation logic multiple outputs in a routing, the attributes of the
•Move logic original entity are automatically copied to each
•Min or Max attribute for selecting incoming newly created entity.
entities as a location rule
The following examples show two methods of
•Routing quantity splitting an incoming batch of material, called
•Routing destination priority BatchA, into one Tote and six EntA’s. The first
•Resource pick up and drop off times method uses a CREATE statement in the opera-
•Location down time logic tion logic to create the new entities, called EntA.
•Debug condition The original entity, BatchA, will have its name
changed to Tote in the routing table. In the sec-
ond example, both of the new entity types are
Memory Allocation for created in the routing table, so no CREATE state-
Attributes ment is needed in the operation logic. In both
cases, the attributes attached to the original
Because locations always exist during a simula- entity, BatchA, are duplicated in both Tote’s and
tion, location attributes always use memory. the EntA’s attributes.
However, attributes for entities are not created
until the first time any of the entity’s attributes
are examined or set. At that time, ProModel allo-
cates enough memory for all of the entity’s
attributes. When the entity exits the system or is
absorbed by another entity through a COMBINE
or JOIN statement, ProModel automatically frees
the memory allocated for its attributes.

Attributes vs. Local Variables


Attributes are primarily useful where the value of
the attribute is assigned in one logic section and
evaluated in another logic section or field, per-
haps at a different location. If, however, an
attribute is assigned a value and evaluated within
the same logic section (e.g., an operation logic), it
ProModel 227
User Guide

Process Logic Notes A general notes field for describing the


attribute. Notes fields contain user comments
only and are never analyzed by ProModel. Click
on the Notes button or double click in the field to
open an edit window for entering detailed notes.

Example of Attributes in Logic


An appliance manufacturer’s model contains an
assembly location to join lids to pots. The pots
are either aluminum or steel and both types of
pots arrive at the same assembly location. If an
aluminum pot arrives at the assembly location, it
must be joined with an aluminum lid. The same
is true for a steel pot and lid. The entities,
steel_lids and alum_lids, are waiting at a queue to
be joined to the pots.
Obviously, one way to model the different pot
types is to use two different entity types. This
example shows how to achieve the same effect
Routing Table using a single entity type (pot) with an attribute
designating whether it is steel or aluminum.
An attribute called “type,” defined in the attribute
edit table, allows the location to tell what type of
pot has arrived at the assembly location. We will
use a value of 1 to represent a steel pot and a
value of 2 to represent an aluminum pot. When a
steel pot enters the system, we assign a value of 1
Attribute Edit Table to the attribute TYPE with the statement
This edit table is used to define entity and loca- TYPE=1. When an aluminum pot enters the sys-
tion attributes. tem, we set its type to 2.

ID The name of the attribute.

Type The type of the attribute, real or integer.

Classification Entity attribute or location


attribute.
228 Chapter 6:
Attributes

At the assembly location, we use the following Process Table


logic:
Entity Location Operation (min)
EntA Loc1 Att1 = 1
WAIT 2 min
EntB Loc2 Att1 = 2
JOIN 1 EntA
EntB Loc3 ...

Routing Table
Blk Output Destination Rule Move Logic
1 EntA Loc2 JOIN 1 MOVE FOR 1
1 EntB Loc3 FIRST 1 MOVE FOR 1
... ... ... ... ...

This logic checks the type of the pot and then In the above example, EntB would have an
joins a lid according to that type. attribute value, Att1, equal to 2 after EntA joined
to EntB.

Attributes and the JOIN State-


Attributes and the GROUP/
ment
UNGROUP Statements
In some cases, one entity joins to another entity
using the JOIN statement (see “Join” on page 504 Suppose several entities, EntA, EntB, and EntC,
for more information). If both entities possess are grouped together and called Batch (see
attributes before they join together, the resulting “Group” on page 493 and “Ungroup” on
joined entity will possess the attribute values of page 568). Each of the original entities have
the entity joined to it. In other words, the entity attributes with values assigned to them before
with the JOIN routing rule is effectively they are grouped. The Batch is processed for 30
destroyed when it gets joined. Consider the fol- minutes, sent to Loc5 and then ungrouped into
lowing diagram in which EntA joins to EntB: the original entities.
The attribute values of the individual entities are
EntA
not transferred to the grouped entity, Batch. In
other words, Att1=0 for the entity, Batch. How-
Join ever, once the entities are ungrouped, they retain

EntB Loc2 EntB

EntA is joined to EntB.

The logic for the above diagram is as follows:


ProModel 229
User Guide

their original attribute values. The following dia- Attributes and the LOAD/
gram graphically shows the concept of grouping.
UNLOAD Statements
EntA
Batch
EntA The LOAD statement loads a specified quantity
Group Ungroup of entities to the current entity. The loaded enti-
EntA
EntB Loc4 Loc5
ties retain their identity for future unloading
EntB
EntBEntC through an UNLOAD statement (see “Load” on
page 509 and “Unload” on page 571 for more
EntC EntC information). When the entities are loaded onto
Three entities are grouped the current entity, the resulting entity retains the
together to form a batch
which is later ungrouped. attribute value of the current entity.
For example, entities called Box are loaded onto
The logic for the diagram is as follows: another entity, Pallet. The Boxes are assigned an
attribute value, Att1=1. Pallets are also assigned
Process Table an attribute value, Att1=2. Once the Boxes are
Entity Location Operation (min) loaded onto the Pallet, the loaded pallet is
EntA Loc1 Att1 = 1 renamed Shipment. The Shipment then has an
EntB Loc2 Att1 = 2 attribute, Att1=2, because it inherits the attribute
EntC Loc3 Att1 = 3 value of the Pallet. However, we then assign an
ALL Loc4 GROUP 3 AS Batch attribute value to Shipment, Att1=3. After the
Batch Loc4 WAIT 30 Boxes are unloaded from the Pallet, the Boxes
Batch Loc5 UNGROUP retain their original attribute value, Att1=1. Now
ALL Loc5 ... the Pallet has a different attribute value, Att1=3,
which was assigned to the renamed entity, Ship-
Routing Table ment. Consider the following diagram and logic
in which two Boxes are loaded onto a Pallet and
Blk Output Destination Rule Move Logic renamed Shipment for the output entity:
1 EntA Loc4 FIRST 1 MOVE FOR 1
1 EntB Loc4 FIRST 1 MOVE FOR 1 Box Box

1 EntC Loc4 FIRST 1 MOVE FOR 1

Load Shipment Unload


1 Batch Loc5 FIRST 1 MOVE FOR 5
Pallet Loc2 Loc4 Pallet
... ... ... ... ... Boxes

Two Boxes are loaded onto a Pallet and renamed Shipment


in the Output. The Boxes are then unloaded from
Please note the Shipment. Shipment is renamed Pallet in the Output.

You can assign an attribute value to a grouped The logic for the diagram is as follows:
entity. However, once the entities are ungrouped,
they retain the attribute values they possessed
before they were grouped.
230 Chapter 6:
Attributes

Process Table bined entity, Box. The following diagram demon-


strates three entities combining into one entity.
Entity Location Operation (min)
Box Loc1 Att1 = 1
WAIT 2 min
EntA
Box Loc2 Att1 = 2
LOAD 2 Combine
Shipment Loc3 Att1 = 3
WAIT 20 EntB Loc4 Box
Shipment Loc4 UNLOAD 2
Box Loc4 WAIT 10
EntC
Routing Table
Blk Output Destination Rule Move Logic
Three entities are combined to
1 Box Loc2 LOAD 1 MOVE FOR 1
form a single entity called Box.
1 Ship- Loc3 FIRST 1 MOVE FOR 3
ment
1 Ship- Loc4 FIRST 1 MOVE FOR .5 The logic for the diagram is as follows:
ment
1 Pallet Loc5 FIRST 1 MOVE FOR 1 Process Table
1 Box Loc5 FIRST 1 MOVE FOR 2
Entity Location Operation (min)
EntA Loc1 Att1 = 1
Attributes and the COMBINE WAIT 2 min
EntB Loc2 Att1 = 3
Statement WAIT 3 min
EntC Loc3 Att1 = 5
Consider the example where several entities are WAIT 6
combined permanently into a single entity, Box ALL Loc4 COMBINE 3
(see “Combine” on page 459 for more informa- ALL Loc5 ...
tion). The combined entity, Box, assumes the
attribute values of the last entity that was com- Routing Table
bined to the single entity. If three entities, EntA,
EntB, and EntC, are combined to form a single Blk Output Destination Rule Move Logic
entity called Box, and EntC was the last entity 1 EntA Loc4 FIRST 1 MOVE FOR 1
that was combined, the Box will have the same 1 EntB Loc4 FIRST 1 MOVE FOR 3
attribute values as EntC. Therefore, if EntC had 1 EntC Loc4 FIRST 1 MOVE FOR .5
an attribute (Att1=5), then Att1=5 for the com- 1 Box Loc5 FIRST 1 MOVE FOR 1
... ... ... ... ...

In the above example, EntC is the last entity to be


combined so the entity Box assumes the attribute
value of EntC, Att1=5.
ProModel 231
User Guide

Variables Variable Edit Table


Variables are of two types: global and local. Glo- This edit table is used to define Variables used
bal variables are place holders defined by the user globally in the model. A description of each field
to represent changing numeric values. Local vari- is given below.
ables are place holders which are available only
within the logic that declared them. Variables can
contain either Real numbers or Integers, includ-
ing element index values, and are typically used
for decision making or recording information. A
Icon This field shows “Yes” if an icon for the
global variable can be referenced anywhere
variable appears on the layout. A variable’s icon
numeric expressions are allowed in a model. If a
looks like a counter and displays the variable’s
variable or attribute is needed only in a single
value.
block of logic, it is easier to define a local vari-
able right inside the logic block. Global variables ID The variable’s name.
are defined in the Variables Editor, accessed from
Type The type of variable, real or integer.
the Build menu. Local variables are defined with
the INT and REAL statements. (See “Local Vari- Initial Value The initial value of the variable to
ables” on page 233 for more information.) be assigned at the start of the simulation. By
default, initial values are 0, but can be changed in
the edit table to whatever value you want. Any
expression can be entered here (including previ-
ously defined variables) except attributes and
system functions.
Stats ProModel collects statistics for each vari-
able on three levels of detail, None, Basic, and
Time Series.
•None No statistics are collected for this
variable during simulation.
•Basic Collects basic statistics such as total
changes, average minutes per change, cur-
rent value, and average value.
•Time Series Collects all the basic statistics
plus the value history based on time or
observations. When you select Time
Series, ProModel collects either time-
How to access the variable edit weighted or observation-based statistics
table:

1. Select Variables (global) from the Build


menu.
232 Chapter 6:
Variables

for the variable depending upon the type Please note


selected.
In order to create plots or histograms for a vari-
able, Time Series stats must be selected in the
variables edit table.

Variable Layout
An icon to show a variable’s value during a simu-
lation may be placed anywhere on the layout. The
Variable Observation window below shows the icons for the variables
Record Value Time in Hours Current and Total at the right side of the screen.
Observation 1 6 1 Each icon has been labeled with a background
Observation 2 5 2 graphic.
Observation 3 6 3
Observation 4 5 4
Total 22 10

Time-Weighted Collects information on the


percentage of time the variable was at a specific
value. As shown in the above table, the average
value of the variable is:
(6 x 1) + (6 x 3) + (5 x 2) + (5 x 4) = 5.4
10
Observation-Based Collects information on
the number of times the variable changed to a
specific value. As shown in the above table, the
value of the variable is 6, then 5, then 6, then 5. How to place an icon for a vari-
The average would simply be: able on the layout:
6+5+6+5 = 5.5
1. Highlight the desired variable in the Vari-
4 able edit table.
Notes A general notes field for describing the 2. Click on the layout where the icon is to
variable. Click on the Notes button or double appear.
click in the field to open an edit window for
3. Size the icon by dragging an edge or cor-
entering detailed notes. ner of the sizing box.
ProModel 233
User Guide

How to remove an icon for a vari- Local Variables


able from the layout:
Local variables function as though they were
temporary attributes defined in a specific logic
1. Double click on the icon.
section which disappear when the logic section is
2. Choose Delete from the resulting menu. finished executing. Local variables are useful for
The icon is removed from the layout, but the test variables in loops and storing locally used,
variable remains in the model. unique values for each entity at the current loca-
tion.
Local variables are used within a block of logic
Editing a Variable’s Icon (i.e., operation logic, subroutines) and are
declared with an INT or REAL statement. Local
A variable’s icon can be customized as necessary variables are only available within the logic in
by simply double clicking on the icon and choos- which they are declared and are not defined in the
ing Edit. A dialog box appears as shown below Variables edit table. A new local variable is cre-
for specifying the characteristics of the variable ated for each entity to encounter an INT or REAL
icon or counter. statement (See “Int” on page 502 and “Real” on
page 539). A local variable is specific to each
entity, in much the same way an attribute is spe-
cific to an entity, except that the local variable is
only available while the entity processes the logic
to declare the local variable. Local variables may
be passed to subroutines as parameters. Local
variables are available to macros.

Example
A plant manufactures valves of 10 different sizes,
How to edit a variable’s icon: such as 10'', 20'', .... All valves are inspected at a
common inspection station and then move to a
dock where they are loaded onto pallets. The pal-
1. Double click on the icon.
lets are designed to hold only a certain size valve.
2. Click on the Digit Color, Frame, or Font but- Therefore, a pallet designed to hold 10'' valves
tons to adjust the respective setting. can only hold 10'' valves, not 20'' valves.
3. Click on OK.
Suppose a Pallet enters a multi-capacity location,
4. All other variable icons that you create Dock. Each Pallet has a different entity attribute,
from now on will retain these modifications. p_type, describing the type of valve it can hold.
Valves are loaded onto the Pallet. The 10” valves
must be loaded onto the pallet designed to hold
the 10” valves. Therefore, the attribute value of
the Valve, v_type, must match the attribute value
of the Pallet, p_type. We can use local variables
234 Chapter 6:
Variables

to accomplish this modeling task. The logic is as enced in the LOAD statement. The pallet
follows where X is a local variable: attribute cannot be directly referenced in the
LOAD statement.
Process Table
If Dock was a single capacity location, using a
Entity Location Operation (min) global variable would work the same as using a
Valve Inspect WAIT 5 local variable. However, because Dock is a
Pallet Dock INT X multi-capacity location, it can load valves onto
X = p_type multiple pallets at the same time. If a global vari-
LOAD 10 IFF able was used instead of a local variable, the glo-
X = v_type
WAIT 10 bal variable would change each time a pallet
entered Dock. If there were two different types
Routing Table of pallets at Dock, there would be only one type
of valve loaded on the pallet because the global
Blk Output Destination Rule Move Logic variable refers to both pallets.
1 Valve Dock LOAD 1 MOVE FOR 2
1 Pallet Delivery FIRST 1 MOVE FOR 8
Suppose, for example, a global variable, type,
signifies the pallet attribute, p_type. We assign
If we had not used local variables, we would need type=p_type at the beginning of the operation
to use the following operation logic for Pallet at logic for location Dock. The first pallet arrives
Dock: and type=3. Therefore, only valves with
v_type=3 are loaded onto the pallet. Another pal-
let enters Dock and type=5. Now only valves
with valve_type=5 are loaded onto both pallets.

Please note

Local variable notes:


1. You may not use the WAIT UNTIL statement
with local variables.
2. The local variable definition only needs to
appear somewhere in the logic before being ref-
erenced. The entity does not need to execute the
As can be seen from the two examples of logic, local variable definition statement (INT, REAL).
the first example is much easier and more straight
forward.
It is important to note that using “LOAD 10 IFF
p_type = v_type” in the operation logic would not
work for the intended purpose. Attributes refer-
enced in IFF conditions always refer to the entity
being loaded. Set the value of a local variable, X,
to the pallet attribute, p_type, so it can be refer-
ProModel 235
User Guide

Arrays Arrays are defined in the Arrays editor accessed


through the Build menu.
An array is a matrix of cells that contain real or
integer values. Each cell in an array works much
like a variable, and a reference to a cell in an
array can be used anywhere a variable can be
used. A one-dimensional array may be thought of
as a single column of values. A two-dimensional
array is like having multiple columns of values
(similar to a spreadsheet).
To reference a cell in a one-dimensional array,
give the name of the array and enclose the cell
number in brackets. For example, if a one-dimen-
sional array containing ten cells is named OpnAr-
ray, to reference the fifth cell you would use
OpnArray[5]. To reference a cell in a two-dimen-
sional array, simply use the name of the array
with the row and column number in brackets. For
example, the statement OpnArray[3,4] references
a cell on the third row in the fourth column.
How to use the arrays editor:
Arrays with more than two dimensions are more
difficult to picture, but work exactly the same as
1. Select Arrays from the Build menu.
one and two dimensional arrays. For example, if
an array has a third dimension, you can reference
any cell simply by adding a comma before the
number of the desired cell in the third dimension. Example Arrays
For example, Tool[3,5,8]. Four and five-dimen- The following examples show how elements are
sional arrays work the same way. The maximum referenced in a one-dimensional array with five
number of dimensions for an array is 20. cells and in a two-dimensional array with fifteen
Array cell values are assigned in the exact same cells.
way as you would assign a value to a variable.
For example, to assign the value 18 to the cell at One-dimensional array
the fifth row and second column in an array OpnArray[5]
named Arr1 you would use the following state-
Cell[1]
ment.
Cell[2]

Assign array cell value Cell[3]


Cell[4]
Arr1[5,2]=18 Cell[5]
236 Chapter 6:
Arrays

Two-dimensional array if your .xls file is in the same directory as your


.mod file, simply enter the name of the .xls file
Tool[3,5]
into this field.
Cell[1,1] Cell[1,2] Cell[1,3] Cell[1,4] Cell[1,5]
See “Export Arrays to Spreadsheets” on page 239
Cell[2,1] Cell[2,2] Cell[2,3] Cell[2,4] Cell[2,5]
for more information on using the Export dialog.
Cell[3,1] Cell[3,2] Cell[3,3] Cell[3,4] Cell[3,5]
Disable: Use this option to have the import file,
export file, or both ignored during simulation
Arrays Edit Table without deleting the name of the import or export
files from the Array record. Choose “None” to
The Arrays edit table is used to define Arrays that use the specified import and export files.
are used in the model. The fields of the Arrays
edit table are explained on the following page. Persist: When running multiple replications
you may choose to keep the values in the array
from one replication to the next or clear (reset)
the array values every replication.
For example, if you run a simulation for three
ID: The name of the array. scenarios with an array that “Keeps” its values,
Dimensions: The size of each dimension of the the data in the array at the end of the first replica-
array in cells. For example, the dimensions of a tion will be kept and used as the second replica-
one-dimensional array of 100 cells is “100.” tion begins. If you are using an Import File, it will
Likewise, a two-dimensional array with 50 rows be used to populate the array for the first replica-
and five columns would have dimensions of tion, but ignored for subsequent replications
“50,5.” The number of rows is first, followed by when “Keep” is selected for the array.
a comma and then the number of columns. An If you choose to “Clear” the array, no array infor-
array cannot have more than 20 dimensions. mation will be kept from one replication to
Type: The type (integer, real, string, or expres- another and the array will be reset with its initial
sion) for all cells in the array. information at the beginning of each replication
(see the next heading “Initializing Arrays”). If
Import File: The name of the spreadsheet from you are using an Import File, it will be used to
which you will populate the array. You may enter populate the array at the beginning of each repli-
either a fully qualified path to your .xls file or a cation when “Clear” is selected for the array.
path relative to your .mod file. For example, if
your .xls file is in the same directory as your Notes A general notes field for entering descrip-
.mod file, simply enter the name of the .xls file tive information about the array. Click the head-
into this field. ing button or double click in this field to open a
larger window for entering notes.
See “Import Data into Arrays” on page 237 for
more information on using the Import dialog. The window pictured above shows how to define
the example arrays that appear on the previous
Export File: The name of the spreadsheet to page.
which you will save the array data. You may
enter either a fully qualified path to your .xls file
or a path relative to your .mod file. For example,
ProModel 237
User Guide

Initializing Arrays tion Logic” on page 180 for a list of logic ele-
ments used to define model parameters.
By default, all cells in an array are initialized to
zero. Initializing cells to some other value should
be done in the initialization logic. A
WHILE...DO loop is useful for initializing array Import Data into Arrays
cell values. The logic below fills a 3 x 5 array (3
rows and 5 columns) with values from an exter- When you import data, from either an external
nal, general read file. Excel spreadsheet or SQL database, into an array,
ProModel loads the data from left to right, top to
bottom. Although there is no limit to the quantity
of values you may use, ProModel supports only
two-dimensional arrays.

Import from an Excel Spreadsheet

This example uses the variables Column and


Row (which may be defined as local variables)
along with two WHILE...DO loops to assign
every cell in the array Inventory_Array a value
from a general inventory file. The logic first sets
the value of the variable Column to one. It then
assigns all the cells in column one a value by
reading a value from the external file and incre-
menting the variable Row. When all the cells in
column one have a value, the logic increments to
the second column and does the inside loop Import File The name of the spreadsheet you will
again. It repeats this loop until each cell in the use. You may enter either a fully qualified path to
array has a value. your .xls file or a path relative to your .mod file.
For example, if your .xls file is in the same direc-
Please note tory as your .mod file, simply enter the name of
the .xls file into this field.
Assigning values to a cell in an array can be done Sheet Name The name of the sheet from which
in any expression or logic field, such as initializa- you will import the array data.
tion and operation logic. However, arrays cannot
be used in logic elements that determine a
model’s structure, such as location capacity. See Please note
“Execution Time of Initialization and Termina-
If your spreadsheet contains only a single data
set, ProModel will automatically load the data
238 Chapter 6:
Arrays

into the array—you do not have to define any cell 6. Enter the Import Start Cell location. The
information unless you wish to limit the contents value in this cell will occupy the first position in
of the array to a portion of the data set. the array.
7. Enter the Import End Cell location. The
Array import requires Microsoft Excel 97 or value in this cell will occupy the last position in
later. the array.
8. Click OK.

Import Start Cell The first piece of data to place


into the array. Import from a SQL Database
Import End Cell The last piece of data to place
into the array.

How to import Excel spreadsheet


data into array:

1. Select Arrays from the Build menu.


2. Select or create the array record (row) you
wish to import.
3. Click the Import File button on the Arrays
dialog.
4. In the File field, enter the name of the
spreadsheet you wish to use (you may also
Connection String The connection string used
browse to select a file).
to connect to the SQL database including all nec-
essary security parameters.
Query or Stored Procedure Enter the SQL
query or the name of the stored procedure to be
executed during the model’s initialization logic to
populate the array.

How to import SQL database data


into array:

1. Select Arrays from the Build menu.


2. Select or create the array record (row) you
wish to import.
5. In the Sheet Name field, enter the name of 3. Click the Import File button on the Arrays
the worksheet that contains the data you dialog.
wish to use.
ProModel 239
User Guide

4. Click on the Database radio button. If you export multiple times to the same spread-
sheet, ProModel will overwrite the spreadsheet
with new data. If you wish to prevent your data
from being overwritten, you can make a backup
of the spreadsheet between each running of the
simulation.
If you are running multiple scenarios and/or rep-
lications, see step 5 below for a description of
how the spreadsheet handles the multiple scenar-
ios/replications.

How to export array data:

1. Select Arrays from the Build menu.


5. In the Connection String field enter the 2. Click the Export File button on the Arrays
necessary string to connect to the desired dialog.
database.
3. Enter the name of the spreadsheet file you
6. In the Query or Stored Procedure field wish to use, or browse to select a file.
enter the SQL query or stored Procedure
name that will be used to populate the array
during the initialization logic.
7. Click OK.

Export Arrays to Spreadsheets

4. Click OK.
5. When the simulation is run, the array data
will be saved to the spreadsheet file. If you
ran multiple scenarios and/or replications,
data from each one will be saved to a sepa-
rate sheet in the spreadsheet file, as seen
below.

The fields in the Export Arrays dialog have simi-


lar functionality as those for the Import Arrays
dialog.
240 Chapter 6:
Arrays

Using Arrays mation and then Arrays. This information


can also be printed.
Using arrays can simplify a model. Suppose you 4. Arrays can be used with the WAIT UNTIL
need to model an assembly line that attaches statement.
components to a computer motherboard. Fur- 5. Statistics are not generated for arrays. How-
thermore, you want to track the usage of compo- ever, if you would like to see the final value
nent parts over time. Without an array, hundreds of an array’s cell, you can use the array
of individual entities of various types would have export feature to export to Excel, you can
to represent hundreds of individual components. place a PAUSE statement in the termination
Keeping track of all the components would be logic and then view the array under the
very complex, not to mention all of the join oper- Information menu, or you could print an
ations and routings for performing the assembly. array’s values or write them to an external
Instead, various cells in a one-dimensional array file as part of the termination logic. If you
could track the number of each type of compo- want more statistical information on a partic-
nent used during the simulation. ular cell, assign the cell to a variable and
An array can do the job more efficiently. The ini- then choose basic or time-series statistics for
tial inventory level for each component could be the variable.
stored in an external file and read into the cells of
the array at the start of the simulation. The first
cell might contain the inventory level of transis-
tors; the second could contain the inventory level
of capacitors and so on. When a motherboard
arrives at the location adding the components,
each cell’s value is decremented according to the
number of that type of component joined to the
motherboard. If each motherboard requires
twelve transistors and five capacitors, then every
time a motherboard arrives at the location, the
array’s first cell is reduced by twelve and the sec-
ond cell is reduced by five. Thus the model
becomes much less complex because it requires
fewer entities and less logic.

Notes on Arrays
1. If a warm-up time is specified, array values
are not reset.
2. Arrays can be nested. For example, if
Arr1[2,3] is equal to three, then the state-
ment Arr2[5,Arr1[2,3]] works exactly like
the statement Arr2[5,3].
3. You can examine the value of a cell in an
array during a simulation by choosing Infor-
ProModel 241
User Guide

Macros Text Any text to be substituted where the macro


name is called. This text may be a complete
A macro is a place holder for an often used expression, an entire logic block, or even part of a
expression, set of statements and functions, or logic block.
any text that might be used in an expression or
Options Allows you to define the macro as a
logic field. A macro can be typed once, and then
run-time interface parameter or select a resource
the macro’s name can be substituted for the text it
group.
represents anywhere in the model and as many
times as necessary. Macros are defined in the The example table above defines three macros.
Macros Editor, accessed from the Build menu. The first macro is simply a numeric constant,
with fpm representing a conversion factor from
miles per hour to feet per minute. If a number is
used in multiple places in a model, then a macro
makes it possible to change that number through-
out the model simply by changing the macro
itself. The second macro, Operation_Time, calcu-
lates the various operation times at different loca-
tions depending on the attributes at the locations.
The last macro, Number_of_AGVs, is a run-time
interface variable used to define the number of
AGVs in the simulation model (see “Run-Time
Interface” on page 242).
A macro is different from a subroutine because a
macro cannot pass or return a value. However,
because it is simply a text replacement, a macro
can reference any expression valid in the expres-
sion or logic field that called the macro. For
instance, the string “the number of entries is”
How to create and edit macros: might be a macro called mac1. This macro by
itself is not a valid expression. However, when
1. Select Macros from the Build Menu. used with the DISPLAY statement in the opera-
tion logic (i.e., DISPLAY mac1), the compiler
will recognize the macro as a string.
Macro Editor A macro may be used in any expression field, but
may only contain a numeric expression (e.g.,
The Macro edit table is used to assign recurring Entries (LOC1), U(5,1), Var1+Att2, etc.). In
text to a reference name. addition, a macro used in an expression field may
not contain multiple lines of text. When used in a
logic field, the macro may include any logic ele-
ment valid in that logic field.
Suppose five different locations use the same
ID A name to identify the macro. lines of code. Instead of entering the same logic
242 Chapter 6:
Macros

five times in five different fields, reference the Please note


following macro by typing the macro ID, Mac1,
in the operation logic of the machine:
Macro notes:
1. A macro may be used only when the elements
contained in the macro are appropriate to the
context from which it was called. This restriction
means that the macro in the previous example is
only valid in operation logic.
2. Because a macro simply substitutes some text
for its name, if a macro represents a statement
block, then it should contain a BEGIN at the
beginning of the block, and an END at the end of
the block. This technique is especially important
when using a macro immediately after a control
statement, such as IF...THEN or WHILE...DO.
For more information, see “Statement Blocks”
on page 436.
Every time the macro is referenced, the logic is
executed. Macros can also be nested within other
macros. This means that a macro can consist of
one or more other macros. Consider the follow- Run-Time Interface
ing Macro edit table:
Defining a run-time interface (RTI) for a macro
allows the user to easily change simulation/model
parameters before the simulation starts. It also
provides an experimental framework for defining
multiple scenarios to be run in a batch (see “Sce-
narios” on page 353). An RTI for a macro allows
a macro’s text to be changed by the user when-
ever a simulation run begins. Since macros are
The macro, Favorite_Quote references other
allowed in any expression, this gives the user
macros, such as Race and Finish. Note that some
flexibility to edit most model parameters every
of the other macros, such as For_Quality, are
time a simulation starts without having to directly
only portions of a complete line of code.
edit the model data.
Although the macro is valid, it will not compile
as a part of macro logic because the create state- The key difference between a macro with an RTI
ment requires an expression and an entity name. and one without is that when a simulation begins,
The line Favorite_Quote in a logic field would be a macro defined with an RTI provides a menu
interpreted as the following line, Create 2 As that allows users to change only the macros you
EntB Take 1 Res1. want them to change. An RTI allows you to
request a variety of information to substitute for a
macro; from simple values (e.g., the initial value
ProModel 243
User Guide

of a location’s capacity) to complex text (e.g., a combinations to find the combination that works
line of logic). You may create RTI parameters best with your model.
using the dialog box below, accessed through the
Numeric Range This specifies the lower and
macros dialog.
upper limits for the parameter if the type is
numeric.

How to define an RTI for a macro:

1. Select Macros from the Build menu.


2. Type the macro name and click on the
Options button.
3. Choose Define from the RTI submenu.
4. Define the Parameter Name and enter the
prompt (optional).
5. Select the parameter type: Unrestricted
Text, Record Range, or Numeric Range.
Parameter Name This text will identify the
parameter represented by this macro. It should 6. If the parameter is numeric, enter the
consist of text that clearly describes the parame- lower value in the From box and the upper
ter to be changed, for example, Operation Time. value in the To box.
The macro name and the parameter name can be 7. Click OK.
different. This provides more flexibility and 8. Enter the default text or numeric value in
allows you to view a more descriptive parameter the Macro Text field.
name when defining scenarios.
9. Use the macro ID in the model (e.g., in
Prompt This text will appear if the user decides operation or resource usage time).
to change the parameter. You should use it to fur-
ther specify the information to be entered, for
example, “Please enter the amount of time the
simulation should run.” Please note
Unrestricted Text This option allows the user to
enter any text, such as the distribution U(8,2). When using a record range, be sure to group all
Note that any text that the user enters will be sub- arrival and shift records. This will allow you to
stituted for the macro name in the model. There- select which series of records to include in the
fore, the text the user enters must be syntactically macro. Note also that when you define an arrival
correct and valid anywhere the macro name or shift RTI, ProModel adds “ARRIVAL_” or
appears. “SHIFT_” to the name to help you identify the
macro more easily.
Record Range Allows you to enable an arrival
or shift record from a range of records. This
allows you to test a variety of shift and arrival
244 Chapter 6:
Macros

Run-Time Interface Example You are now able to change the operation time at
Lathe_1 using the Model Parameters option in
Suppose you build a factory model and determine the Simulation menu. For more information on
the first lathe, Lathe_1, is a bottleneck. The changing model parameters and defining scenar-
model results indicate the throughput is lower
ios, see “Model Parameters & Scenarios” on
than expected. You decide to perform several
page 352.
what-if scenarios with the model by changing the
operation time of Lathe_1. Instead of changing
the operation time at Lathe_1 within the Process Please note
edit table, it is easier to define a macro with an
RTI. This technique allows the model user to For more information concerning the differences
easily see the effect of installing a faster lathe between macros and subroutines, see “Macros”
without ever editing the model itself. The fol- on page 241 and “Subroutines” on page 246.
lowing example represents the dialog used to
define the RTI for the macro where the operation
time is a numeric value between 12 and 20:
Resource Grouping
Resource grouping allows you to define specific
groups of resources rather than define each unit
separately. For example, suppose you need a spe-
cific technician to perform an operation. If the
technician is not available, you may use either
another technician or one of two qualified opera-
tors to perform the operation. Rather than define
each qualified operator as a separate resource,
you may define a macro that includes them.

How to define a resource group


After defining the RTI for the macro, substitute
1. Select Macros from the more elements
the macro, Oper_Time, for the operation time in
section of Build menu.
the operation logic in the Process edit table for
Lathe_1 as shown below: 2. Define a macro ID and enter a list of all
resources you wish to include as part of the
resource group.

Please note

When you create a list of resources, separate


each resource using AND or OR (e.g., Tech_1
AND Tech_2 OR Tech_3 AND Tech_4).
ProModel 245
User Guide

3. Click on the Options button and select


Resource Group from the submenu.

At the end of the simulation, ProModel creates a


statistical report containing information collected
for each resource included in the resource group,
as well as the entire group. This will allow you to
track individual, as well as group performance.
For information on statistics and how to graph the
results, see “Reports and Graphs” on page 373.
246 Chapter 6:
Subroutines

Subroutines Subroutines are defined in the Subroutines Editor


which is accessed from the Build Menu.
A subroutine is a user-defined command that can
be called to perform a block of logic and option-
ally return a value. Subroutines may have param-
eters (local variables) which act as variables local
to the subroutine and that take on the values of
arguments (i.e., numeric expressions) passed to
the subroutine.
ProModel handles subroutines in three ways.
First, a subroutine may be processed by the call-
ing logic as though the subroutine is part of the
calling logic. This way is the most commonly
used, and is done by simply referencing the sub-
routine by name in some logic or expression.
Second, a subroutine may be processed indepen-
dently of the calling logic so the calling logic
continues without waiting for the subroutine to
finish. This method requires an ACTIVATE
statement followed by the name of the subroutine
(see “Activate” on page 441), or you may use the How to create and edit subrou-
Interact Menu (see “Run-Time Interact Menu” on tines:
page 368). Third, ProModel allows subroutines
written in an external programming language to 1. Select Subroutines from the Build menu.
be called through the XSUB() function.

Subroutine Editor
The Subroutines edit table consists of several
fields which identify the components of a subrou-
tine. Each of these fields is described below.

ID A name that identifies the subroutine.

Type The type of numeric value returned by the


subroutine can be Real, Integer, None, or Interac-
tive. Use Real and Integer if the subroutine
returns a number and None when no return value
is expected, as is often the case in initialization or
ProModel 247
User Guide

termination logic. Subroutines of type Interactive fore, independent subroutines can run in parallel
are identical to subroutines of type None, except with the logic that called them.
that interactive subroutines are also accessible for
4. Independent subroutines called with the ACTI-
activation by the user through the run-time menu.
VATE statement cannot contain entity-specific or
Interactive subroutines are displayed in the Inter-
location-specific system functions.
act menu during runtime. For more information
on Interactive subroutines, see the discussion
later in this section.
Parameters Arguments passed to the subrou- Subroutine Format
tine get assigned to local variables called parame-
ters. Items passed to a subroutine as arguments A subroutine may be named any unique, valid
can have different names than the parameters that name. The general format for calling a subrou-
receive them. Parameters can be real or integer. tine is as follows:
The first parameter receives the first argument, SubroutineName(arg1, arg2,....,argn)
the second parameter receives the second argu-
ment, and so on.
Subroutine call
Logic One or more statements to be executed
whenever the subroutine is called. Statements in GetOpTime(3,7)
subroutines must be valid in the logic that calls
DoInitialization()
the subroutine. Subroutine logic may contain a
RETURN statement with a value to be returned
from the subroutine. (See “Return” on page 549
for the correct syntax and an example of this Please note
statement.)
Subroutine format notes:
Please note
1. If no arguments are specified, open and closed
Subroutine editor notes: parentheses are still required.
1. If the subroutine is of type Integer and the 2. Statements in subroutines must be valid in the
return value is a real number, the return value logic that called the subroutine. For example, if a
will be truncated unless the ROUND() function is subroutine is called from the operation logic, the
used (e.g., RETURN ROUND(<numeric expres- subroutine may contain only those statements
sion>)). valid in the operation logic. Subroutines called
2. If you do not want a stand-alone subroutine from an ACTIVATE statement or from the Inter-
referenced in operation logic to be treated as an act Menu at run-time can have any general logic
implicit wait statement, define the subroutine as statements, including WAIT.
type None.
3. A subroutine may be used in any logic field. In
3. When using the ACTIVATE statement to call a addition, a subroutine may be used in any expres-
subroutine, the calling logic continues without sion field, provided that the RETURN statement
waiting for the called subroutine to finish. There- is used to return a value to the expression field.
248 Chapter 6:
Subroutines

Expression fields include the Qty Each column of routing are shown below. The subroutine itself is
the Arrivals edit table and the routing rule for shown later in this section.
processing.

4. If a subroutine does not return an expression


with the RETURN statement, a value of zero will
be returned for subroutines of type Real and Inte-
ger. No value will be returned for a subroutine of
type None or Interactive.

Subroutine Example
Suppose that you build a Copy Center model and
it is necessary to define the processing time at the
operator assisted machines as a function of two
parameters: Order Quantity and Quality Level. In
addition, you would also like to write the pro-
cessing times at each copy machine to a file
called Report.Dat. Subroutines Edit Table
Because the operation logic is identical at both The Subroutine edit table lists the name of the
machines, a subroutine may be used to execute subroutine, the return type, the parameters to be
the group of statements required. Notice in the passed to the subroutine, and the logic. Notice
Process Logic pictured next that Subroutine that the subroutine ID corresponds to the subrou-
OrderTime is called, and each order's Quantity, tine name called in the processing logic above
Quality, and Ticket attributes are passed as argu- and that the subroutine has four parameters (P1,
ments. The third argument is simply a constant P2, P3 and P4) corresponding to the four argu-
value, 1 or 2, which represents the location from ments passed to the subroutine. Also note that
which the subroutine was called (i.e., 1 if called the subroutine is of type real since the return
from OpAsist1 or 2 if called from OpAsist2). The value will be a real valued processing time.
operation logic, which calls the subroutine, and

Subroutine Parameters
The subroutine parameters, M (for mean) and SD
(for standard deviation), are defined by clicking
on the Parameters heading button. These values
are unique to each inspection location, and are
ProModel 249
User Guide

passed to the subroutine as parameters of the nor- upon the quality level desired. Operation time is
mally distributed inspection time. determined by simply multiplying the process
time by the scale factor.
Once the order time has been determined, this
value is written to the file Report.Dat. Included
with the operation time is the job number and the
location at which the job was processed.
The last line of the logic returns the order time
value to the processing logic.

Subroutine Logic Interactive Subroutines


The final step in defining the subroutine is to
Interactive subroutines are subroutines activated
specify the logic to be processed when the sub-
by the user anytime during run-time by choosing
routine is called. In this example, the logic should
the subroutine from the Interact menu. The name
include a processing time, a procedure to write
appearing in the Interact menu is either the sub-
the values of the attributes to a file, and a routing
routine name or if a string is entered as a com-
decision based on the values of the two attributes.
ment statement at the beginning of the subroutine
The logic window is accessed by clicking on the
logic, the string is used as the name (e.g., #
Logic heading button.
“Arrival Frequency”). Interactive subroutines
allow the user to interact with the simulation dur-
ing run-time. Subroutines are defined as type
Interactive in the Subroutine edit table. Normally,
subroutines are activated by entities. However,
interactive subroutines can be user-activated in
addition to being entity-activated. Interactive
subroutines can be used for:
•changing model parameters during a simula-
tion run
•changing routings
•calculating and reporting user-defined statis-
tics
Suppose you want to interactively change the
The first line of the logic calculates the order time arrival frequency for a certain entity, customers,
as a function of the Quantity and Quality during run time. Define a variable, Var1, and
attributes passed as parameters. These parame- assign it an initial value to be used for the initial
ters are used as arguments in two user defined arrival frequency. Enter Var1 in the arrival fre-
functions, ProcTime() and QualFactor(). Func- quency field for the entity customers. Create a
tion ProcTime() returns a time value based on the
number of copies in the order, while function
QualFactor() returns a scale factor depending
250 Chapter 6:
Subroutines

subroutine of type Interactive and enter the fol- knowledge of your external programming lan-
lowing logic: guage (C, C++, Pascal) to use external subrou-
tines. In addition, you must also have a good
Windows platform knowledge, specifically with
respect to creating DLLs in your language.
Because it is a 32-bit program, ProModel can
load and call only 32-bit DLLs, and requires that
you use a 32-bit Windows compiler.
For more information about this feature, you can
load, study and run XSUB.MOD in the reference
model directory (also see “Xsub()” on page 583).
This model uses XSUB.DLL, found in the MOD-
ELS directory. The source code and make files
for XSUB.DLL (XSUB.CPP, XSUB.MAK,
During run-time, you can then change the arrival XSUB.IDE) are also included in the MODELS
frequency for the customers by choosing the Cus- directory. Some general explanation is contained
tomer Arrivals from the Interact menu (see “Run- in the comments of this source code.
Time Interact Menu” on page 368).
Due to the complexities of Windows program-
ming and the variety of uses for this advanced
Please note feature, PROMODEL Corporation can only pro-
vide minimal support for this feature. Many ques-
Interactive subroutines may also be called from tions regarding Windows programming and other
any logic or expression where no return value is programming languages cannot be handled
required. See “Statements and Functions” on through our customer support department. Please
page 439 for more information. consult your language programming manuals,
language customer service centers, Microsoft,
and other resources to resolve these types of
problems.
External Subroutines
There may be some cases where you need to per- Subroutines vs. Macros
form actions ProModel is not capable of doing.
You may need extended capabilities with more Although subroutines and macros work similarly,
sophisticated commands. ProModel allows you they have subtle differences. Any logic may use
to interface with external subroutines located in both macros and subroutines. The main differ-
thirty-two bit Windows DLL files you have cre- ence is in the way they are used. Only subrou-
ated. This feature could be useful for doing tines can be used when you need to pass
sophisticated file I/O, performing statistical anal- arguments, get a return value, or activate the
ysis, making your simulation interactive, or help- independent execution of logic. Only macros can
ing with other simulation needs. be used when defining run-time interface param-
eters.
Because of the intricacies of the Windows devel-
opment environment, you must have a sound
ProModel 251
User Guide

Macros may be used in any expression field, but


the macro may only contain an expression (i.e.,
Entries(LOC1), U(5,1)). When a macro is used in
a logic field, the macro may include any logic
element valid in that logic field.
Subroutines may also be used in an expression
field provided that the RETURN statement is
used to return a value to the expression field.
When a subroutine is used to represent one or
more logic statements, the subroutine may only
include statements valid for the particular con-
text.
252 Chapter 6:
Arrival Cycles

Arrival Cycles Arrival Cycles Edit Table


An arrival cycle is a pattern of individual arrivals Arrival cycles are defined by entering the proper
which occurs over a certain time period. Some data into the Cycle edit table. The fields of the
examples of arrival cycles that exhibit a pattern Cycle edit table are explained below.
are the arrival of customers to a store and the
arrival of delivery trucks to a truck dock. At the
beginning of the day, arrivals may be sparse; but
as the day progresses, they build up to one or
more peak periods and then taper off. While the
ID The cycle name.
total quantity that arrives during a given cycle
may vary, the pattern or distribution of arrivals Qty / % Select either Percent or Quantity as the
for each cycle is assumed to be the same. basis for the total number of arrivals per cycle
occurrence.
Arrival Cycles are defined in the Arrival Cycles
edit table, accessed through the Build Menu. Cumulative... Select Yes to specify the % or
Qty values in a cumulative format. Select No to
specify these fields in a non-cumulative format.
Table... Click on this button (or double click in
this field) to open an edit table for specifying the
cycle parameters.

Arrival Cycles Example


Suppose we are modeling the operations of a
bank (or any service or manufacturing system)
and we need to specify a pattern for customer
arrivals. From past data, we know that customers
arrive throughout the day (9:00 AM to 5:00 PM)
according to the following approximate percent-
ages.
From Before Percent

How to edit arrival cycles: 9:00 AM 10:30 AM 0


10:30 AM 11:30 AM 10
1. Choose More Elements from the Build 11:30 AM 1:00 PM 25
Menu. 1:00 PM 4:00 PM 55
4:00 PM 5:00 PM 70
2. Choose Arrival Cycles.

Defining the Arrival Cycle


The arrival cycle will be named Bank_Arrivals in
the Arrival Cycles edit table. Because the data is
ProModel 253
User Guide

expressed in terms of percentages, we select Per- record for the bank example appears below. (See
cent as the basis for the cycle. Also, the percent- “Arrivals” on page 163 for more information.)
age values are not cumulative so we specify No
in the Cumulative field.

The total number of customers per day is nor-


mally distributed with a mean of 1000 and a stan-
Next, we click on the Table heading button to dard deviation of 35. The number of occurrences
open another edit table for entering the cycle of the cycle is 20, representing 20 working days
data. (1 month) of time. The frequency in this case
refers to the period of the cycle or the time
between the start of one cycle and the start of the
next cycle, which is every 24 hours. Make sure
the arrival frequency is defined with the same
time unit as the arrival cycle.
Without an arrival cycle, all the Customers for a
single day would arrive at the start of the simula-
tion. An arrival cycle will divide the quantity
specified in the Qty each... field into various
In this example, even though the percentages of sized groups arriving throughout the day.
customers that have arrived are not cumulative,
the time is always cumulative. Therefore, the To assign the arrival cycle to the arrival record,
table reads as follows: ten percent of the daily click on the Qty each... heading button to open a
customers arrive in the first 1.5 hours, fifteen per- dialog with the names of all defined cycles.
cent of the daily customers arrive between hour
1.5 and 2.5, and so on. The arrivals are randomly
distributed according to a Uniform distribution
during the time interval in which they arrive.
The arrival cycle is now defined and can be
assigned to an arrivals record in the Arrivals edit
table.

Assigning Arrivals to the Arrival


Cycle
After an arrival cycle has been defined, assign an
arrival record to the arrival cycle in the Arrivals
edit table for the arriving customers. The arrival
Click on the entry Bank_Arrivals and select OK.
The Qty each... field now includes the cycle.
254 Chapter 6:
Arrival Cycles

Cumulative Cycle Tables the number of arrivals to arrive within each time
interval.
In the previous example, percentages were
expressed non cumulatively. This same data Example 1
could have been expressed cumulatively as fol-
Suppose that in the bank example we knew that
lows:
for each cycle period, the number of customers to
Before Percent arrive during each time interval within the cycle
period is as follows:
9:00 am 0
10:30 am 10 From Before Number
11:30 am 25
9:00 AM 10:30 AM 100
1:00 pm 55
10:30 AM 11:30 AM 150
4:00 pm 70
11:30 AM 1:00 PM 300
5:00 pm 100
1:00 PM 4:00 PM 150
The data is now expressed cumulatively and 4:00 PM 5:00 PM 300
could be entered in the cycle table as follows.
With the data in this format, we specify the
Arrival cycle by choosing Qty in the “Qty/Per-
cent” field and complete the cycle table as fol-
lows:

To specify cycles in cumulative form, simply


choose Yes in the Cumulative... field of the
Arrival Cycles edit table.
This table could also be specified cumulatively
Please note by choosing Yes in the Cumulative... field and
entering the quantity values in a cumulative for-
Time values remain cumulative regardless of the mat.
form of the percentages.
Please note

When specifying Arrival cycles by quantity, the


Arrival Cycles by Quantity value entered in the “Quantity each” field of the
The previous example was based on the assump- Arrivals edit table changes meaning. Instead of
tion that a certain percentage of arrivals came the total arrivals per cycle, it represents a factor
within a specified time interval. An alternate by which all entries in the cycle table will be mul-
method of specifying an arrival cycle is to specify tiplied. This field may be any valid expression
which evaluates to a number. This allows the
ProModel 255
User Guide

same arrival cycle to be entered for more than


one arrival record that has a different factor
applied to it. For example, you might want to
define a different factor to a customer arrival
pattern depending on the day of the week. In this
case, define an arrival record for each day with a
frequency of one week.

Example 2
Suppose we wish to see the effect on the Bank
example if the number of arrivals is increased by
50%. The relative quantities per time interval
remain the same but now 50% more customers
arrive each day. Using the data from the previous
example, we enter the same values in the Arrival
cycle quantity fields, but specify a value of 1.5 in
the quantity field of the Arrival edit table.

The result is an arrival schedule with the follow-


ing parameters.
From Before Quantity
9:00 AM 10:30 AM 150
10:30 AM 11:30 AM 225
11:30 AM 1:00 PM 450
1:00 PM 4:00 PM 225
4:00 PM 5:00 PM 450
256 Chapter 6:
Table Functions

Table Functions ear interpolation. The following two examples


show how to specify a linear function (where
Table functions provide an easy and convenient only two reference values are needed to define
way to retrieve a value based on an argument the entire function) and a nonlinear function
(i.e., some other value) that is passed to the table. (where more than two reference values need to be
Table functions specify a relationship between an specified).
independent value and a dependent value. All
table functions are defined in the Table Functions Example 1
editor which is accessed from the Build menu.
The example below shows a relationship that
exists between the time required to process an
order and the number of entities in the order. As
the number of entities increases, the time
required to process the order also increases. The
relationship in this case is linear, meaning the
processing time is directly proportional to the
number of copies in the order. Because of the lin-
ear relationship, only the two endpoints need to
be entered in the function table. (The function
table for this example is given in the discussion
on the Table Function editor.)
Function: ProcTime()
3

Processing
time (min)
How to access the function table 1

editor:
0
0 100 200 300
1. Select More Elements from the Build menu.
Order size
2. Select Table Functions.
Example 2
In this example the relationship between the inde-
Table Functions Editor pendent value and the dependent value is nonlin-
ear and inversely proportional. In addition,
Table functions are defined by the user and return interpolation is required to determine the depen-
a dependent (or look-up) value based on the inde- dent value if the independent value passed to the
pendent (or reference) value passed as the func- function lies between the independent values
tion argument. Independent values must be
entered in ascending order. If the independent
value passed to a table function falls between two
independent values, a dependent value for the
unspecified reference value is calculated by lin-
ProModel 257
User Guide

given explicitly in the table function. The tables for the two example functions are
given below.
Function: QualFactor()
3.0

2.0
Processing
Time Factor
1.4

1.1
1.0

1 2 3 4 5

Quality Level

In the example above the dependent value repre-


sents a factor which is to be multiplied by the
processing time required to complete an order.
Each order is assigned a number, 1 through 5,
according to the level of quality desired. In this The independent and dependent values allow any
case, a lower number represents higher quality. general expression such as numbers, variables,
(The function table for this example is listed in and math functions. The fields are evaluated only
the following explanation of the Function Table at translation, and cannot vary during the simula-
Editor.) tion.
When calling a user-defined table function, if the
Table Function Edit Table independent value is out of range, then the table
function will return a zero for the dependent
The Function Table Editor is where all function value. Consider the following function table,
tables are created and edited. The fields of the Operation_Time:
Function Table editor are defined below.

ID The name of the function table.

Table... Click on this heading button to open a If the function were called with the command
table for defining the independent and dependent “Operation_Time(5),” the independent value
values of the function. passed to the table function Operation_Time
would be five. But five is beyond the limits of the
table, so the dependent value returned will be
zero. Likewise, if the independent value is 1, the
dependent value returned will be zero. However,
if 2.7 is entered as the independent value, Pro-
258 Chapter 6:
Table Functions

Model will interpolate and return a value between


30 and 50.
ProModel 259
User Guide

User Defined Distributions section). The data is entered into the User Distri-
bution edit table. The User Distribution edit
Occasionally, none of ProModel’s built-in distri- table’s fields are described below.
butions can adequately represent a data set. In
these cases, the user may define a User Distribu-
tion to represent the data set. User Distributions
specify the parameters of user-defined (empiri-
cal), discrete, or continuous probability distribu-
tions.
ID The name of the distribution. When referenc-
ing distribution tables (in the operation logic, for
example) the open and closed parentheses after
the distribution name must be used, such as
Dist1(), OpTime().
Type... Discrete or Continuous depending on the
number of possible outcomes.
Cumulative... Yes or No depending on whether
the distribution is to be specified in cumulative or
non-cumulative format.
Table... Click on this button (or double click in
the field) to open an edit table for defining the
parameters of the distribution. Once a distribu-
tion has been defined, the field changes from
“Undefined” to “Defined.”
The combination of Discrete and Continuous dis-
How to create and edit user distri- tributions, along with the ability to express either
butions: in cumulative or non-cumulative terms, creates
four possible formats for specifying distributions.
1. Choose More Elements from the Build
The remainder of this section gives examples and
menu. procedures for specifying each of these distribu-
tion types.
2. Choose User Distributions.

Discrete Distributions
User Distribution Edit Table Discrete distributions are characterized by a finite
set of outcomes, together with the probability of
A user-defined distribution is a table of empiri- obtaining each outcome. In the following exam-
cally gathered data. User distributions may be ple, there are three possible outcomes for the
either continuous or discrete, and may be cumu- group size: 30% of the time the group size will be
lative or non-cumulative (more information con-
cerning these options is found later in this
260 Chapter 6:
User Defined Distributions

10, 60% of the time the group size will be 20, and Process Table
10% of the time the group size will be 30.
Entity Location Operation (min)
Discrete Distribution EntA Loc1 GROUP Dist() AS Batch
Batch Loc1 WAIT 10 min
P 1.00
r .90
Routing Table
o .80
b .70
Blk Output Destination Rule Move Logic
a .60
b .50
i .40 1 Batch Loc2 FIRST 1 MOVE FOR
l .30 5
i .20
t .10 ProModel provides the flexibility to specify dis-
y 0 crete distributions according to a probability
10 20 30
mass function or a cumulative distribution func-
Group size tion. Select Yes or No in the Cumulative field of
the Distribution edit table and fill in the table
One way to represent a discrete distribution is by according to the probability mass function or the
its probability mass function, listing the possible cumulative distribution function. The following
outcomes together with the probability of observ- tables show the discrete distribution example
ing each outcome. A probability mass function defined in both formats.
for the example above could be expressed as fol-
lows (with G representing the group size). Discrete (probability mass
G 10 20 30 function)
P(G) .30 .60 .10
An alternate way to represent a distribution is
through a cumulative distribution function, list-
ing each possible outcome together with the
probability that the observed outcome will be less
than or equal to the specified outcome. A cumu-
lative distribution function for the example above
could be expressed as follows.
Discrete (cumulative distribution
G 10 20 30 function)
P(G) .30 .90 1.0
In the next example, the number of parts are
grouped into a batch according to a user distribu-
tion.
ProModel 261
User Guide

Continuous Distributions P(3.5 <= X < 5.0) = 0.40


P(5.0 <= X < 6.0) = 0.30
Continuous distributions are characterized by an P(6.0 <= X <= 8.0) = 0.10
infinite number of possible outcomes, together P(8.0 < X) = 0.00
with the probability of observing a range of these The following table represents the p.d.f. for this
outcomes. In the following example, there are example.
an infinite number of possible operation times
between the values 2.0 minutes and 8.0 minutes.
Twenty percent of the time the operation will take
from 2.0 to 3.5 minutes, 40% of the time the
operation will take from 3.5 to 5.0 minutes, 30%
of the time the operation will take from 5.0 to 6.0
minutes, and 10% of the time the operation will
take from 6.0 minutes to 8.0 minutes.

As with a discrete distribution, a cumulative dis-


Continuous Distribution tribution function for a continuous distribution
specifies the probability that an observed value
P 100
r 90 c.d.f. will be less than or equal to a specified value. A
o 80 c.d.f. for the example distribution is as follows
b 70 (where x represents the return value).
a 60
b 50 x 2.0 3.5 5.0 6.0 8.0
p.d.f.
i 40
P(X <= x) 0 .20 .60 .90 1.0
l 30
i 20 The following table represents this c.d.f.
t 10
y 0

1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0

Processing time (min)

As with a discrete distribution, a continuous dis-


tribution can be defined in two ways. A probabil-
ity density function lists each range of values
along with the probability that an observed value
will fall within that range. Each of the values
within the range has an equal chance of being
observed, hence the piece-wise linearity of the
c.d.f. within each range of values. A probability
density function for the example above is
expressed as follows.
P(0.0 <= X < 2.0) = 0.00
P(2.0 <= X < 3.5) = 0.20
262 Chapter 6:
External Files

External Files during the simulation. Each of these fields is


defined below.
External files may be used during the simulation
to read data into the simulation or write data as
output from the simulation. Files can also be used
to specify such things as operation times, arrival
schedules, shift schedules, and external subrou-
tines. All external files used with a model must ID An alias to be used in the model for referenc-
be listed in the External Files Editor which is ing the file. Note that this ID does not have to be
accessed from the Build menu. the same as the file name.
Type Click on this heading button to display the
following menu. The six file types are discussed
in the remainder of this section.

File Name... The actual file name, including the


path. Press the heading button or double click in
this field to browse for a specific file.
Prompt A prompt to be displayed at run time in
the event that the specified file cannot be opened.
How to define external files:
Notes... A general notes field for entering
1. Select More Elements from the Build menu. descriptive information about the file. Click the
heading button or double click in this field to
2. Select External Files. open a larger window for entering notes.

File Types
External Files Editor
External files may be defined as one of several
The External Files Editor consists of an edit table types depending upon the purpose of the file.
with fields specifying the external files to be used
General Read File
A General Read file contains numeric values read
into a model using a READ statement. Values
must be separated by a space, comma, or end of
ProModel 263
User Guide

line. Any non-numeric data will be automatically entered in the cell where the names intersect. An
skipped when obtaining the next numeric value example of a spreadsheet file is shown next.
(See “Read” on page 537 for correct syntax and
examples). For example, if you specify a normal
distribution such as N(5,1) in the General Read
file, ProModel will not return a numeric value
following the distribution. Instead, it will read in
the first value, 5, and the next value, 1.
A General Read file must be an ASCII file. Data
created in a spreadsheet must be saved as a text
file.

To use the value stored in an Entity-Location file


General Write File
as an operation time, call out the file identifier in
A General Write file is used for writing text the operation logic as shown in the following
strings and numeric values using the WRITE and example. (In this example, “SvcTms” is the File
WRITELINE statements. Text strings are ID of the desired Entity-Location file.)
enclosed in quotes when written to the file, with
commas automatically appended to strings. This Please note
enables the files to be read into spreadsheet pro-
grams like Excel or Lotus 1-2-3 for custom view-
If the .wk1 or .xls file contains more than one
ing, editing, and graphing. Write files may also
sheet of data, only the first sheet will be read in.
be written to using the XWRITE statement which
gives the modeler full control over output and
formatting. (See “Write” on page 581, “Write-
Line” on page 582, and “Xwrite” on page 585 for
Process Table
correct syntax and examples.)
Entity Location Operation (min)
If you write to an external file during multiple
EntA Loc1 WAIT SvcTms()
replications or a single, independent run, the data
will be appended to the data from the previous
Routing Table
replication. However, if the RESET statement is
used, the data is overwritten for each replication. Blk Output Destination Rule Move Logic
1 EntA Loc2 FIRST 1 MOVE FOR 5
Entity-Location File By specifying SvcTms() with no arguments in the
An Entity-Location file (or expression file) is a parentheses, a value is returned from the Entity-
spreadsheet (.wk1 or .xls format only) file con- Location File “SvcTms” for the current entity at
taining numeric expressions listed by entity and the current location, i.e., EntA at Loc1. You may
location name. Entity names should appear also return the value stored in any other cell of an
across the top row, beginning in column 2, while Entity-Location file by explicitly specifying the
location names should be entered down the first entity and location names in the parentheses, e.g.,
column, beginning on row 2. A numeric expres- SvcTms(EntB, Loc1) or SvcTms(EntC, Loc2).
sion for each Entity-Location combination is
264 Chapter 6:
External Files

Arrivals File Qty, Time, Number, and Frequency columns


in a spreadsheet.
An Arrivals file is a spreadsheet (.wk1 or .xls for-
When defining an External Arrivals File, you do
mat only) file containing arrival information nor-
not need to define arrivals in the Arrivals edit
mally specified in the Arrival Editor. One or
table. If several entities are scheduled to
more arrival files may be defined and referenced arrive at the same time, entities arrive in the
in the External Files Editor. Arrival files are system according to the order in which they
automatically read in following the reading of the appear in the arrival list. However, when
Arrival Editor data. The column entries must be there is more than one occurrence for the
as follows: arrival record, the next entity will not arrive
until the frequency has elapsed. Meanwhile,
Column Data
other entities listed below the record may be
A Entity name allowed to arrive.

B Location name
C Quantity per arrival
Please note
D Time of first arrival
E Number of arrivals If the time of the first arrival is zero and there is
only one arrival of some quantity, you do not
F Frequency of arrivals
need to complete additional cells. Likewise, if
G through... Attribute assignments there is only one arrival at a time other than zero,
you do not need to fill in additional cells after the
Columns A through F may have any heading
Time entry.
desired as long as the data is of the proper type.
If attributes are to be assigned, columns G and
higher should have headings that match the
names of the attributes being assigned. The fol-
lowing example illustrates these points.
Please note

If the .wk1 or .xls file contains more than one


Example sheet of data, only the first sheet will be read in.

Shift File
A shift file record is automatically created in the
External Files Editor when you assign a shift to a
The values in the spreadsheet cells must be a location or resource. If shifts have been assigned,
numeric expression as opposed to a formula the name(s) of the shift file(s) will be automati-
commonly used in spreadsheets. For exam- cally created in the External Files Editor. If no
ple, if cell E4 in the spreadsheet above was path is listed for the shift file, ProModel will
actually a formula generating the value 100,
search in the default models directory.
the value ProModel generates is zero. Pro-
Model only recognizes expressions for the
ProModel 265
User Guide

Please note RDW Binary Output Reports &


Graphs saved for a
Creating a shift file record in the External Files model
Editor should not be done. It is done automati- SED* Binary Seed File used to store
cally through the shift assignment. seed values for each rep-
lication
SFT Binary Shift created when
DLL File defining a shift in the
shift editor
A DLL file is needed when using external sub-
routines through the XSUB() function. See “Sub- TRC* ASCII Trace File
routines” on page 246 for more information. GLB Binary Graphic Library Files
GBM Binary Graphic Bit Map file
Excel File created when loading a
An Excel file is automatically created in the model file.
External Files Editor when you assign a shift to
an array import file. If the external file has been Open Files
assigned, the name(s) of the Excel file(s) will be
automatically created in the External Files Editor. Depending on the model specifications, there
If no path is listed for the Excel file, ProModel may be some occasions where several files need
will search in the default models directory. to be open simultaneously to execute the model.
There is a feature which allows the user to access
up to 255 open files at any time. The [General]
Other External Files section of the promod.ini file contains the fol-
In addition to allowing the user to define external lowing statement: OpenFiles=n where n is the
files, ProModel creates other external files. Pro- number of files between 20 and 255. The default
Model automatically creates and/or opens files is 40. To change the number of available open
depending on the specifications in the model. files, simply edit thepromod.ini file such that
Below is a description of the different files Pro- OpenFiles equals the desired number and then
Model creates (* indicates files that remain open restart ProModel. There is also the option to close
while the model is running): files using the CLOSE statement (see “Close” on
page 458).
Extension Type Description
CSV ASCII Export Data (comma
delimited)
MOD* Binary Model File
RDB Binary Output Database (basic
statistics)
RDT Binary Output Time Series data
266 Chapter 6:
Streams

Streams Streams Edit Table


A stream is a sequence of independently cycling, The Streams Editor consists of an edit table with
random numbers. Streams are used in conjunc- three fields. The fields of the Streams edit table
tion with distributions. Up to 100 streams may be are explained below.
used in a model. A stream generates random
numbers between 0 and 1, which in turn are used
to sample from selected distributions. By default,
all streams use seed value #1 and are not reset
between replications if multiple replications are
run. To assign a different starting seed value to a
stream or to cause the seed value to be reset to the
initial seed value between replications, use the Stream # The stream number (1 to 100). This
Streams Editor. The Streams Editor is accessed number identifies each stream.
from the Build menu as shown below.
Seed # The seed value (1 to 100). Streams hav-
ing the same seed value generate the same
sequence of random numbers. For more informa-
tion on seed values, see “Using Random Number
Streams” on page 266.
Reset... Set this field to YES if you want the
stream to be reset to the initial seed value for each
model replication. Set this field to NO if you
want the stream to continue where it left off for
subsequent replications.
The Streams Editor pictured above shows five
defined streams. Each stream has a unique seed
value. Streams 1, 2 and 5 will be reset for each
replication, while streams 3 and 4 will continue
where they left off for subsequent replications.

Using Random Number


How to access the streams editor:
Streams
1. Select More Elements from the Build menu. One of the most valuable characteristics of simu-
2. Select Streams from the submenu. lation is the ability to replicate and isolate proba-
bilistic functions and activities within a system
for specific study. In the real world, events tend
to occur randomly, according to a certain statisti-
cal pattern or distribution. To help you model this
randomness, ProModel uses distributions.
ProModel 267
User Guide

When you include a distribution (e.g., Normal, activities will not affect the sample values gener-
Beta, and Gamma) in your model, ProModel uses ated from the arrival distribution).
a random number generator to produce a set
sequence or stream of numbers between 0 and 1 Please note
(0 <= x < 1) to use in the distribution. Before it
can select any numbers, however, ProModel
The random number generator is a prime modu-
requires an initial seed value to identify the point
lus multiplicative linear congruential generator.
in the stream at which to begin. Once you specify
The C code implementation for most of the ran-
a seed value, ProModel “shifts” the random num-
dom variates was written by Stephen Vincent and
ber selection (in increments of 100,000 numbers)
based on the algorithms described by Law and
by that number of positions and starts sampling
Kelton (see the “Bibliography” on page 606).
values. Since there is only one random number
stream, this will ensure that the selected values
do not overlap. ProModel includes 100 seed val-
ues, and each seed produces a unique set of ran-
dom numbers. If you do not specify an initial Stream Example
seed value, ProModel will use the stream number The following example shows one reason why
as the seed value (i.e., stream 3 uses seed 3). multiple streams are useful.
Random Number Stream Two machines, Mach1 and Mach2, are to go
100,000 random numbers down approximately every 4 hours for servicing.
To model this, the frequency or time between
}

failures is defined by a normal distribution with a


mean value of 240 minutes and a standard devia-
Seed 1 Seed 2 Seed 3 tion of 15 minutes, N(240,15). (For more infor-
mation on distributions, see “User Defined
When you use a specific seed value (e.g., 17), Distributions” on page 259). The machines will
ProModel produces a unique sequence of num- go down for 10 minutes. Because no stream is
bers to use each time you apply that seed value. specified in the normal distribution, ProModel
This allows you to maintain the consistency of uses stream number one to generate sample val-
some model elements and permit other elements ues for both machines. So if the next two num-
to vary. (To do this, specify one random number bers in stream number one result in sample values
stream for the set of activities you wish to main- of 218.37 and 264.69, Mach1 will receive 218.37
tain constant, and another random number stream and Mach2 will receive 264.69. Therefore, the
for all other sets of activities.) In fact, because two machines will go down at different times,
each seed value produces the same sequence of Mach1 after almost four hours and Mach2 after
values every time, completely independent model somewhat more than four hours.
functions must use their own streams. For exam-
ple, Arrival distributions specified in the Arrival Suppose, however, that the resource to service
Module should have a random number stream the machines must service them both at the same
used nowhere else in the model. This will prevent time. The machines would have to go down at
activities that sample random stream values from the same time. Stream number one will not bring
inadvertently altering the arrival pattern (i.e., the them down at the same time because stream num-
ber one sends the machine’s distributions two dif-
268 Chapter 6:
Streams

ferent numbers. Using two streams (in the Note that if a third machine were to use one of the
example below numbered ten and eleven) with streams, for example if Mach3 were to use stream
the same initial seed will ensure that the eleven, the two machines would no longer go
machines receive the same random number every down together. Mach1 would use the first value
time. The two streams have the same starting from stream ten; Mach2 would use the first value
seed value so they will produce exactly the same from stream eleven; and Mach3 would use the
sequence of random numbers. The first number second value from stream eleven. The first time
of stream ten will be exactly the same as the first Mach1 and Mach2 went down, they would go
number of stream eleven; the second numbers down at the same time because the first number
will be the same; indeed, every number will be in streams ten and eleven is the same. But there-
the same. after they would not. The second time they
would go down at different times because Mach1
The Streams window below shows streams ten
would receive the second value from stream ten,
and eleven with the same seed values.
Mach2 would receive the third value from stream
eleven, and Mach3 would receive the fourth
value from stream eleven.

Please note
By assigning stream ten to Mach1 and eleven to Stream notes:
Mach2, both machines will go down at exactly
the same time. 1. If a stream value is not specified for a distribu-
tion, the stream is assumed to be stream one.
Stream #1 does not automatically reset after each
replication.

2. The stream parameter is always the last


parameter specified in a distribution unless a
location parameter is also specified. See “Distri-
bution Functions” on page 437 for details.

This first clock downtime for each machine will


occur at the beginning of the simulation. After
that, the first downtime occurs at 218.37 minutes
for both machines. Defining different seed val-
ues for the streams would produce different
sequences and therefore different downtimes.
Using the same stream number for the clock
downtimes would also produce different values.
But two different streams with the same seed
value will bring both machines down at the same
times.
ProModel 269
User Guide

Material Handling Systems if bridge B tries to move to a node N that lies very
close to the left end of the envelope, and bridge A
The following section provides advanced tech- is to the left of bridge B, a run-time error will
niques to model Material Handling Systems. occur if the space remaining between node N and
bridge A is less than the bridge separation dis-
tance. To avoid such problems, define the crane
Crane Systems envelope wide enough to allow sufficient space
To implement cranes in ProModel, the Path Net- beyond any serviceable nodes.
work module allows three types of networks:
non-passing, passing, and cranes. From this mod- Crane Priorities, Preemption &
ule, you can lay down and orient a crane enve- Bridge Bump-Away
lope, define the bridge separation distance, and
define the graphics for rails and bridges in the ProModel associates two types of priorities with
bay. Additionally, the Resources module allows the operation of crane resources: (1) resource
you to define bridge and hoist speeds, and hoist request priority, and (2) crane move priority.
graphics for one or more cranes operating in the Resource Request Priorities Used to
same bay. resolve any conflicts that arise between multiple
tasks requesting the same resource at the same
time. This priority scale follows the rules that
Creating Multi-Bridge Crane
regular dynamic resource priorities do (i.e., 10
Systems levels of 100). Use this priority to preempt crane
Modeling multiple cranes operating in the same resources from lower priority tasks to serve
bay has never been so easy! higher priority tasks.
User defined bridge separation distances, exten- Crane Move Priorities Used for multiple
sions to regular dynamic resource usage state- bridges operating in the same envelope, to decide
ments, and a set of new priority rules have been which crane bridge has priority over another
introduced in order to let you manage conflicting while moving. This priority scale is also in the
movements in the same envelope zone. range [0...999], but does not have any preemption
Crane Envelope The parallelogram-shaped levels. Any higher value has overriding priority
area represented by a crane type path network, over a lower value. You can assign priorities to
bounded by two rails and the lines connecting the crane resources temporarily, on a task-by-task
endpoints of those rails. The lines connecting the basis via extra parameters in the GET, JOINTLY
endpoints of rails are effectively the two extreme GET, USE, and MOVE WITH statements.
positions of the center-lines of any bridges oper- Therefore, you can only use priorities for cranes
ating within the envelope. ProModel uses the end that are either moving to respond to a resource
of one of the rails as the envelope origin to serve request or moving to deliver a part. Moving to
as a reference point for all logical distance calcu- downtime, off-shift or break nodes has the
lations within the envelope. default (0) move priority. Moving to park has the
Bridge Separation The minimum distance lowest move priority (the same as an idle crane).
you want to maintain between center-lines of two You can further sub-divide crane move priorities
neighboring bridges. This distance is not related into three categories:
to how close a bridge center-line can get to one of •Task Move Priority Assigned by the
the extreme endpoints of the envelope. Caution: modeler via parameters passed to resource
270 Chapter 6:
Material Handling Systems

usage statements (GET, MOVE WITH, etc.). •Crane Graphics In the Resources module,
Used as the base move priority for travel to the selected graphic represents the hoist. You
pick up and travel to deposit. may define, edit, and size multiple graphics
•Claim Priority If crane A’s bridge is for the same hoist the same way you create
headed towards its ultimate (task) destina- multiple graphics for other resources.
tion, the claim priority is equal to the task •Units & Multiple Cranes Crane resources
move priority of crane A. Otherwise (if crane limit you to one unit. When you define mul-
A’s bridge is moving under the influence of tiple cranes operating in the same envelope,
crane B), claim priority is equal to the claim define them as separate resources—each
priority of the claim inducer bridge (bridge assigned to the same crane network.
B). •Downtimes In addition to the clock and
•Effective Claim Priority Applies only usage downtimes you may define for cranes,
to envelopes with three or more cranes. If you can assign cranes to shifts in the same
multiple crane bridges are moving in the way you assign other resources. When a
same direction with overlapping claims, the crane goes down, it moves to the downtime
effective claim priority of any bridge in the node specified in the DTs dialog. If you do
overlapping region is the maximum of all the not associate a node with the crane down-
claim priorities of those bridges. time, the crane goes down at its current posi-
tion.
Crane Operations •Specifications Cranes require separate val-
ues for bridge and hoist speed and accelera-
The operation of cranes in ProModel has been
tion. You must enter Speed (Empty), Speed
designed to closely resemble the real-life opera-
(Full), and Accelerate and Decelerate values
tion of cranes. Under most circumstances, you
for bridge movement along the rails and
need not be concerned with how ProModel han-
hoist movement across the bridge. In each
dles crane movements. The following operational
field, the format is Bridge Value, Hoist
specifications are intended as a reference only
Value (e.g., if you enter “150, 20” in the
and are not essential in order to model cranes
speed field, the bridge moves at 150 and the
using ProModel.
hoist at 20 feet per minute). Although you
must enter speed values, acceleration and
Crane Animation deceleration values are optional (if you leave
During a simulation run, entities picked up by a them blank, ProModel assumes the crane to
crane appear graphically on the entity spot for the have an infinite acceleration and deceleration
hoist. The hoist graphic appears above the entity capability. (Hint: this provides better run-
graphic and the bridge of the crane appears on top time computational efficiency).
of the entity and hoist graphics. •Crane Searches Define Work and Park
searches in the same way you define
searches for other resources.
Crane Specifications
•Node Logic Define entry and exit logic for
When you define cranes for your model, consider cranes at nodes in the same way you define
the following: node logic for other resources.
ProModel 271
User Guide

Handling Zone Claims In Multi- Case 1


Crane Envelopes
If a crane is in one of the following states, it will
be referred to as “immovable at a node.” No other
crane will be able to push this crane away,
regardless of its priority.
•Picking up or depositing an entity.
•Being used at a location. A B' B A'
•Having a downtime, off-shift time, or break
time. Bridge B starts moving first, from [B] to [B'] and
claims the zone [B'~B] in the ← direction. Bridge
Extended Movement Zone The net zone
A wants to move from [A] to [A'] in the opposite
that it needs to traverse ± bridge separation dis-
direction. If the move priority of A is less than or
tance (± depending on the direction that it will
equal to the move priority of B, we let bridge A
go).
start moving towards [B' - bridge separation]
Before starting to move, any <crane A> attempts only. If A has a higher move priority, B is
to claim its extended movement zone (zone A) stopped immediately and sent back to [A' +
for the entire movement duration. bridge separation]. In the latter case, A may
ProModel can reject the claim attempt of crane A queue up behind B.
for zone A for several reasons:
•If there are immovable cranes in zone A Case 2
•If there are any cranes in zone A with move
priorities greater than or equal to1 the move
priority of crane A, if they are going in the
opposite direction, or they are going in the
same direction with conflicting destina-
tions.2

Case Examples of Zone Claims A B A' B'


The following cases illustrate specific situations Bridge B starts moving first, from [B] to [B'],
with examples: then bridge A wants to move from [A] to [A'].
Bridge B claims zone [B~B'] in the → direction.
Since bridge A wants to move in the same direc-
1. Special case treatment not “>=” but strictly “>” if all tion and their destinations do not conflict (assum-
of the following conditions hold: (1) if the bridge ing that B'-A' is greater than the bridge
that A is trying to oppose is stationary, (2) unless separation), bridge A moves all the way. If B has
crane A is trying to move just to park (if A has a task
or if it is trying to move out of the way of another already cleared the zone [A~(A' + bridge separa-
bridge), (3) if the one A is trying to push is idle or its tion)] by the time A wants to start moving, A
bridge was blocked before reaching its destination. moves independently. Otherwise, A may queue
2. Destination Conflict Test: If sign (B' - A' - up behind B. (If the distance B'-A' is less than the
sign(existing claim)*(bridge_separation)) = sign given bridge separation and the move priority of
(existing claim) then not conflicting. A is greater than the move priority of B, bridge A
272 Chapter 6:
Material Handling Systems

will push away bridge B, and crane B will not direction with a higher move priority than both of
respond to the request while passing over [B']). the other bridges. ProModel interrupts bridges A
and B and lays a new course for both bridge B
Case 3 [B'' = C' - bridge separation] and bridge A [A'' =
B'' - bridge separation]. In this case, B may queue
up behind A, and C behind B.

Case 5

A B A'' B' A'

Bridge B starts moving first, from [B] to [B'],


then bridge A wants to move from [A] to [A'].
A B C
Bridge B claims the zone [B~B'] in the → direc-
tion. Since bridge A wants to move in the same B'
direction and their destinations do conflict, prior- A'
( (
ities will be considered. If the move priority of A C'
is not greater1 than the move priority of B, bridge
A starts moving towards [A'' = B' - bridge separa- This is an extension of Case 4. Assume that the
tion] only. Otherwise, A will push B away, and B move priorities for crane bridges A, B, and C are
will not respond to the request while passing over 3, 1, and 2 respectively. When bridge C tries to
[B']. In both cases, A may queue up behind B. claim zone [C'~C] in the ← direction, it faces an
“effective priority” of 3 and is unsuccessful in its
Case 4 original claim, so it starts moving towards [B' +
bridge separation]. Bridge B goes outside the
effect of bridge A in the second part of its move-
ment, [(A' + bridge separation) ~ B'] at time t0.
However, bridge C does not override bridge B
and start moving towards [A' + 2 * bridge separa-
tion] until the first “reclaim trigger event2” after
time t0.
If ProModel rejects the original claim of crane A,
A B C' A' B' C
crane A claims the largest available portion of the
Bridge B starts moving first from [B] to [B'], original zone, and implements the procedure
claiming zone [B~B'] in the → direction. Bridge described below to move its bridge through the
A wants to move from A to A' and starts moving subset zone. The crane waits until a “reclaim trig-
behind B. Bridge C wants to move in the opposite ger event” occurs, and then repeats the claim
attempt to let its bridge traverse the remaining

1. In this special case, bridge A’s claim priority com-


petes with the task move priority of bridge B. 2. Discussed later in this section.
ProModel 273
User Guide

distance. The hoist, on the other hand, moves are compared with the a, s, and d of the leading
directly to its relative destination on the bridge, bridge (L). If all of the following conditions are
independent of the bridge zone claim results. In true, bridge F moves independent of bridge L.
case of bridge interruption, the hoist does not
stop moving. Therefore, for any particular crane Conditions for No Queue-Up
move, the combined trajectory of hoist and 1. aF <= aL The following bridge cannot accel-
bridge may change depending on zone availabil- erate faster than the leading bridge.
ity at the time of the claim. 2. sF <= sL The following bridge’s maximum
If crane A accepts the claim attempt, all cranes speed cannot be greater than the leading
inside the claimed extended zone bump away at bridge.
the time of the claim. This claim also prevents 3. dF >= dL The following bridge can deceler-
any other cranes with move priority less than or
ate equally or faster than the leading bridge.
equal to the move priority of A from entering the
claimed zone in conflicting directions. If any one of the above conditions fail, bridge F
moves at the same speed as bridge L, maintaining
the distance that existed between them at the time
Zone Claim Notes F started moving.

Zone Claim Notes


Please note
1. A crane cannot do a pickup or deposit while
another crane pushes it away, even though
Condition 3 usually creates a deviation from the
the request might be on its way.
typical real life situation, but it is necessary to
2. When you interrupt a crane bridge and ask it
avoid time consuming computations that can
to make a new claim (or update its claim), its
adversely affect the run-time performance of
speed is reset to zero and ProModel re-evalu-
models with crane resources. To avoid unrealis-
ates its mobility characteristics. ProModel
tic queue-ups when modeling multiple bridges
recalculates the move time with these new
with different mobility characteristics in the same
parameters. This rule also applies to bridges
crane envelope, make either no bridge decelera-
that are moving when a bridge behind them
tion entry (infinite deceleration), or use the same
with a higher priority and a conflicting desti-
overall average deceleration rate for all crane
nation makes a new claim to go in the same
bridges in that envelope.
direction. See Case 3 (when priority of A is
greater than priority of B) for an example.

Treatment of Potential Queue-Up


Situations
For all bridge claims with the potential of having
two bridges queue up after one another (men-
tioned in the example cases above), the following
rule applies:
The acceleration rate (a), maximum speed (s) and
deceleration rate (d) of the following bridge (F)
274 Chapter 6:
Material Handling Systems

Bridge Motion after Queue-Up another). Independent of its current position and
movement direction, the crane still has the capa-
distance
traveled bility to respond to resource requests coming
by bridges from anywhere in the envelope. If allocated to a
task with high enough travel-to-pickup move pri-
Bridge
Separation ority, the crane may reverse direction or termi-
nate its move early, bumping other crane bridges
B on its way.
Qup
Whenever a crane bridge changes its claim, it
distance checks to see if there are any bridges behind it
A
that are going in the same direction with overlap-
time
B accel. B max. B decel. ping claims. If there are any, the newly claiming
A mimics B A goes with B’s crane triggers a chain queue-up re-evaluation
(former) average speed
towards the following cranes. The following
If bridge A is queued up behind bridge B, bridge example demonstrates this case.
A goes through up to two stages of movement.
The first stage occurs throughout the movement B''2 B''1
of bridge B. Bridge A mimics bridge B’s motion,
maintaining its original distance from bridge B.
The second stage occurs after bridge B stops.
Bridge A continues traveling with the former
average speed of bridge B until it reaches its des-
tination.
A B A' B'
If bridge B is moving to accomplish its own task
(as opposed to just moving out of A’s zone), then At time t0, crane A is called for a task and starts
bridge A might stop earlier, without going moving to [A'] with move priority mpA, pushing
through the second stage. In this case, the total idle bridge B to [B' = A' + bridge separation]. At
traveling time for bridge A is calculated as if time t1 (bridge A and B are still moving), a
bridge A were traveling with bridge B’s average request for crane B comes from [B" (1 or 2)] with
speed. move priority mpB. Bridge B is allocated to the
When a bridge arrives at its destination, it trig- task immediately since it is available at the time.
gers a chain queue-up re-evaluation that pene- If the new destination of bridge B is [B"1], the
trates the crane envelope in the opposite direction new claim of bridge B does not conflict with the
that the stopped bridge was moving. This chain existing claim of bridge A, so the move priorities
re-evaluation handles cases in which there were are not significant. ProModel re-evaluates all
multiple bridges queued up behind the stopped motion values for crane B (acceleration, deceler-
bridge. ation, empty and full speeds for both the hoist
and the bridge) and crane B starts moving to its
Moving Idle Cranes can Respond to new destination. If the bridge motion expressions
Resource Requests for crane B have changed their values between t0
and t1, the queue-up status between bridges B
A crane can be idle and moving at the same time and A (and any others behind A) may change, so
(moving to park or being pushed away by
ProModel 275
User Guide

ProModel performs a chain queue-up re-evalua- In the first six cases, a crane changes state from
tion. “immovable” to a finite move priority. The sev-
If the new destination of bridge B is [B"2], the enth case may change the “effective move priori-
new claim of bridge B conflicts with the existing ties” of other cranes. ProModel requires the
claim of bridge A, so ProModel considers the eighth case for envelopes with three or more
move priorities: bridges with overlapping operation zones: It
•If mpB <= mpA, then B does not have ensures that two or more unsatisfied, opposing
enough priority to stop A before A gets to its claims with the same move priority will resolve
destination. Bridge B will have to go on their conflict based on earlier task assignment
moving until [B'] and wait until it can claim time instead of locking up the crane system.
the region [B"2 ~ B'] in the → direction. Hence, all of the above cases create a potential to
realize previously unsuccessful claims.
•If mpB > mpA, then B does have overriding
priority over A. Bridge B stops bridge A at The reclaim trigger events for each envelope
[B"2 - bridge separation], starting a chain echo in the trace together with the triggering rea-
move termination which tells bridge A to son for model debugging purposes. The results of
stop any other followers with the appropriate reclaim attempts for each crane resource and its
bridge separation distances between each. In effects on other crane resources in the same enve-
this case, bridge A (and other possible fol- lope are also displayed.
lowers) will have to wait until the next
reclaim trigger event to attempt reaching Other Crane Operation Rules
their original destinations. 1. If an entity currently possesses a crane, all
move logic for the entity must use a MOVE
Reclaim Trigger Events WITH statement (until you free the crane)—
Unsuccessful claims of all crane bridges in an the entity cannot move without the crane.
envelope are re-attempted when a reclaim trigger 2. Although an entity cannot possess more than
event occurs in that envelope. The following one crane at a time, the entity may possess a
events are considered to be triggers for claim re- crane and a resource of another type.
evaluations in the related envelope: 3. If a crane is in one of the following states,
another crane cannot push it away.
•The duration of a USE <crane> statement
a) Picking up or depositing an entity.
expires.
b) Being used at a location.
•You free a crane using the FREE statement.
•A crane completes a deposit (unless you
c) Experiencing a downtime, off-shift
time, or break time.
specify the KEEP option).
4. The hoist moves to its relative destination on
•A crane completes a pickup and the crane is
the bridge independent of the bridge zone
about to start moving to deliver.
claim attempt outcomes. The hoist does not
•A crane downtime ends.
stop for bridge interruption so, for any par-
•You preempt a crane resource.
ticular crane move, the combined trajectory
•A bridge arrives at its task destination of hoist and bridge may change depending
(before elapsing pickup/deposit time). on zone availability at the time of the claim.
•There are no moving task bridges left (for 5. When a crane pushes away an idle crane, the
envelopes with 3 or more cranes only). idle crane comes under the control of the
276 Chapter 6:
Material Handling Systems

work and park search lists defined for the minimum distance to the hoist and then the
node closest to its bridge. order of appearance in the nodes edit table.
If the hoist is still moving at the time the
Crane Operations Notes bridge stops, the registry procedure executes
by considering the destination of the hoist.
ProModel allows definition of resource points for
crane resources on envelope nodes to achieve the
three dimensional look for “lowering the hoist” Please note
upon arriving at a node. Since cranes can only
have one unit, only the first resource point for Only users can instruct cranes to park on nodes.
any crane at any node is meaningful. Also, Pro- However, as a result of bump-aways, a crane can
Model only uses the resource points if the crane end up standing still at positions that do not nec-
resource arrives at the node to perform a task. In essarily correspond to nodes. In any case, Pro-
other words, the hoist does not lower if the crane Model registers an idle stationary crane to a
arrives at a node to park or just ends up there after node and places it under the control of work and
a bump. park searches of that node.
In general, once ProModel makes a match
between a resource request and a resource of the
requested type, the resource is allocated for that
job and the simulation engine does not go on Cranes follow the same rules as regular dynamic
looking for other matches that can actually reach resources while searching for assignments
to the requester sooner. Since cranes blocking (returning to preempted jobs, exclusive & non-
other crane resources is a common phenomenon exclusive work searches, requests coming from
in multi-bridge envelopes, extra care should be nodes not included in the work search, etc.).
used while designing the active operation zones There are some additional steps to handle interac-
and resource request logic in such models. tions with other cranes operating in the same
envelope. Upon completion of its task, a crane
goes through these extra steps:
Nodes, Work, and Park Searches
1. Lets other cranes in the same envelope re-
ProModel “registers” all cranes to their home attempt their claims, possibly generating an
nodes at the beginning of a simulation run. When induced claim on its bridge.
a crane starts to move, ProModel erases registra- 2. Even though other cranes might bump away
tion, and re-establishes it when the crane stops, its bridge, the crane executes the work search
according to the following rules: defined for it at its current node. (If requests
1. Cranes assigned to a task (pick up, drop off, are waiting at different locations of the same
down) remain unregistered throughout inter- work search record, ProModel takes the
mediate stops and register to the task desti- resource request priorities into consider-
nation node upon arrival. ation.)
2. Idle cranes that are bumped away, or cranes 3. If it was not allocated to a task or bumped
that were unsuccessful in claiming the entire away, the crane executes the park search
bridge region to go parking register them- defined for it at its current node, or attempts
selves to the node closest to their bridge going home if return home is checked.
when they stop. If there are multiple nodes
closest to the bridge, ties are broken by the
ProModel 277
User Guide

Please note time a crane would have taken to get to its desti-
nation if it were the only bridge in its envelope.
A crane bridge moving to park cannot bump
other idle bridges out of its way. In such a case, Please note
the bridge goes as far as it can towards its desti-
nation and registers itself to the closest node The first component of gross travel time (moving
upon stopping (using the rules discussed above). towards the actual destination) may be greater
than the net travel time, since it may include
going back towards the destination after another
crane bumps it away. It may also involve devia-
Crane Resource Statistics tions from the bridge’s original motion values
The statistics reported for cranes are very similar when queued-up behind slower bridges.
to regular dynamic resource statistics. The fol-
lowing definitions highlight the differences in the
interpretations.
In Use Includes the following states: Special Case—Blocked Time
•Stationary use at a location via GET, Accrual for Moving to Park
JOINTLY GET, and USE statements. Remember that an idle crane registers itself to the
•Gross move time while delivering — actu- closest node whenever its bridge stops. If the
ally carrying an entity from one location to hoist is still moving, ProModel uses the destina-
another. tion of the hoist to break any ties between nodes
•Pickup and deposit times before and after the that are equidistant to the bridge position.
move to deliver. In case of an unsuccessful claim attempt to move
Travel To Use Is the gross move time to start to park, the crane bridge needs to stop before
being used — traveling to respond to a request reaching the instructed park node, and ProModel
for either stationary use or use to pickup and ignores the hoist travel for statistics purposes.
deliver an entity to another location. The “travel to park” statistic terminates at the
Gross Move Time May include the follow- time the bridge stops, not when bridge or hoist
ing times: stop time reaches its maximum value. In this
case, the net moving time to park computes as the
•Moving towards the actual destination.
bridge move time from current position to the
•Possibly moving away from the destination,
endpoint of the claim, using original bridge
when bumped away by a higher priority
motion values for the crane resource. Blocked
bridge in the same envelope.
time can accrue if this bridge queues up after a
•Stationary blockage times during which the slower bridge that is already moving (remember
bridge is waiting for a higher move priority that a crane moving to park cannot bump another
bridge in the same envelope to lower its crane).
move priority.
Blocked Time Tally blocked time by sub-
tracting the internally computed, hypothetical net Conveyors
traveling time from the actual incurred gross A conveyor is any moving track, belt, chain, or
move time. ProModel defines the net time as the roller which transports parts from one location to
278 Chapter 6:
Material Handling Systems

another. This section will help you better under- Conveyor Types
stand conveyors and how to model them using You may represent conveyors as one continuous
ProModel. span of randomly spaced parts (e.g., belt con-
ProModel determines the capacity of a conveyor veyor) or as intermittent spaces or carriers
by its speed and load spacing rather than a stated located at fixed intervals (i.e., trolley conveyor).
capacity. Specifically, capacity is a function of
the minimum allowable interload spacing on a
conveyor (which is the length of a queue position
Accumulating & Non-Accumulating
in the case of accumulation conveyors) and its In addition to continuous or intermittent convey-
length. It may be desirable, however, to impose a ors, some conveyors permit loads to accumulate
limit to the number of loads permitted on the con- or queue if conditions impede forward motion
veyor at any one time. progress. ProModel refers to these conveyors as
Conveyors usually do not pick up and drop off accumulating conveyors. Other conveyors pro-
loads as in the case of lift trucks, AGVS, and vide no queuing and halt all activity if forward
cranes. You must place loads onto and remove progress of the leading part or load stops. Pro-
them from conveyors. Model refers to these types of conveyors as non-
accumulating conveyors.
Depending on the nature of the conveyor and its
operation, modeling a conveyor can be straight- The following figure illustrates how entities can
forward or complex. For single conveyor sec- continue to accumulate after the leading load
tions, modeling is very simple. Conveyor stops on an accumulating conveyor. How a non-
networks, on the other hand, give rise to several accumulating conveyor can accept one additional
complexities (recirculating, merging, etc.) that entity if the leading entity stops, however it is
become more difficult to model. unable to advance any further since the entire
conveyor stopped.
Conveyor Simulation Benefits
Simulating your conveyor system will help you
address design and operational issues, answering
the following questions:
•What is the minimum conveyor speed that
still meets rate requirements?
•What is the load rate capacity of the con-
veyor?
•What is the load delivery time for different
activity levels?
•How much queuing do accumulation con-
veyors need?
•How many carriers do trolley or power-and- The different combination of continuous or inter-
free conveyors need? mittent conveyors and transport or accumulation
•What is the optimum number of pallets to conveyors result in the following four different
maximize productivity in a recirculating kinds of conveyors that ProModel is capable of
environment? modeling.
ProModel 279
User Guide

Single Section Conveyors


Conveyors used for simple queuing or buffering
Type Example often consist of a single stretch or section of con-
veyor. Loads enter at one end and exit at the other
Continuous, Non- Belt, chain end. Load spacing is generally random. These
Accumulating types of conveyors are generally quite easy to
model. The modeler merely defines the length
Intermittent, Non- Tray, trolley
and speed and states whether the conveyor is an
Accumulating
accumulating or non-accumulating conveyor.
Continuous, Roller
Accumulating Conveyor Networks
Intermittent, Power-and-free, A conveyor network consists of two or more sec-
Accumulating towline tions that are connected together to enable part
merging, diverting, and recirculating. In such
For best results, you should model carriers or part instances, a conveyor may have one or more
holding devices on intermittent, accumulating entry points and one or more exit points. Further-
conveyors as though the devices were entities more, part entry and exit may not always happen
waiting for you to load and unload other entities at the beginning or end of the conveyor. In Pro-
to and from them (see "Load" is ProModel User Model, loads or entities must always enter at the
Guide) as you move parts. An alternative but beginning and exit at the end of conveyor net-
similar method is to use a JOIN statement and works. To model merging or branching in the
then specify two routings when you want the part middle of the conveyor, break the conveyor into
removed (see Example E in "Modeling Conveyor smaller sections.
Systems" in the next section). Yet another way of
modeling parts with carriers on a conveyor, is to Modeling Conveyor Systems
simply set a part attribute equal to a particular
A conveyor system is a combination of one or
value if the part is on a carrier. If the part is not on
more types of conveyors which interact to move
a carrier, assign a different value to the same
parts. Such a system may include branching,
attribute. Operations may occur at either the
merging, and looping. The process of dividing
beginning or end of a conveyor as shown here.
conveyors into logical sections is fairly straight-
forward if you understand what constitutes a con-
veyor section.
For modeling purposes, a conveyor section is
defined as any independently driven segment of
conveyor of the same type. Conveyor turns are
irrelevant as long as the conveyor remains
unchanged and both are actuated or stopped
together.
In ProModel, an accumulating conveyor section
must end at the point where parts are permitted to
280 Chapter 6:
Material Handling Systems

accumulate or back up. This generally occurs at Some users prefer to display the grid and select
a transfer point or process location. the Snap to Grid option in the View menu. Others
will leave Snap to Grid off. You may find that the
Modeling Conveyors precise positioning of a conveyor is easier when
you turn Snap to Grid off and set the Zoom suffi-
While not essential, it is visually more effective
ciently high. When you need to change the type
to draw conveyors to scale on the layout. You
of conveyor to accumulating, double-click on the
should also size entities traveling on a conveyor
conveyor icon in the layout, select the Conveyor
to scale when you create them. Setting the length
Options button, and select Accumulating from
of a conveyor different from the scale length or
the Conveyor Options dialog.
conveying entities that have a length or width that
The following six examples show various ways
is not to scale will result in the appearance of load
of modeling conveyors:
overlapping or excessive load spacing. The
underlying logic and simulation is still valid, Example A Routes one or more parts from
regardless of whether the graphics are to scale. one or more non-conveyor locations to any con-
Loads may appear to jump when transferring veyor location.
from the end of one conveyor to another con-
veyor. This is a result of ProModel not allowing
a load to transfer onto a conveyor until there is
room for the load. Including a transfer time in
the Move Logic column of the Routing edit table
can result in smooth movement between con-
veyor sections. Defining a variable called
Transfer_Time and using that variable rather than E l A
Example B Routes one or more parts from
entering the actual time allows for experimenta-
any conveyor location to one or more non-con-
tion with various conveyor speeds with minimal
veyor locations.
modeling changes.
Clock precision is also important. Note that
movement of one foot at 40 feet per minute will
take .025 minutes. If clock precision is set at .1
minutes, truncation will result in zero time. Make
sure you check your model for the correct clock Example C Routes single parts from any con-
precision. For most conveyors, a clock precision veyor location to one or more other conveyor
of .001 is adequate. ProModel may require locations.
greater precision for high-speed conveyors where
speeds are greater than 200 fpm.
Modeling conveyors in ProModel is quite easy.
Select the conveyor icon in the Locations Graph-
ics window and place the beginning and ending
points of the conveyor using the mouse. Clicking
the right mouse button will end the conveyor.
Left-click to add or delete a joint and allow the
conveyor to bend.
ProModel 281
User Guide

Example D Joins one or more parts from one Automated Guided Vehicle
or more non-conveyor locations to another trans-
porting entity (e.g., a pallet or carrier) at any con- Systems
veyor location. An automated guided vehicle system (AGVS) is
a path network along which computer-controlled,
driverless vehicles transport loads. One of the
modeling requirements of AGVS is to accurately
describe the method for controlling traffic. You
can usually accomplish this in one of two ways:
•Zone blocking
•On-board vehicle sensing
Zone Blocking The most common method of
traffic control, it involves placing control points
along the guide path. Each point usually allows
Example E Accumulates multiple parts com-
only one vehicle to access it any one time, thus
ing from one or more non-conveyor locations
blocking any other vehicle from entering any seg-
onto a conveyor location.
ment of the path connected to that point. Once the
vehicle leaves a control point to travel to the next
point on the path, any vehicle waiting for access
to the freed control point can resume travel.
On-Board Vehicle Sensing Works by hav-
ing a sensor on-board the vehicle that detects the
presence of a vehicle ahead of it and stops until it
Example F Splits parts at a conveyor location detects that the vehicle ahead of it has moved.
sending one or more parts to a non-conveyor
location and a single part (e.g., a pallet or carrier) Modeling AGVS
onto the conveyor. Modeling an AGVS is very similar to modeling
an industrial vehicle, such as a lift truck, (which it
is in a sense) except you control the operation
more and there exists less freedom of movement.
Paths are generally unidirectional and do not
allow vehicle passing.
One of the challenges in modeling an AGVS is
finding the shortest routes between any two stops
in a system. ProModel provides built-in capabil-
ity to automatically determine the shortest routes
between points in a complex network.
282 Chapter 6:
Material Handling Systems

AGVS Simulation Benefits •What is the best path layout to minimize


travel time?
Simulating your AGV system will help you
address design and operational issues, answering •Where are the potential bottleneck areas?
the following questions: •How many people do you need to meet
activity requirements?
•What is the best path layout to minimize
travel time? •What are the best shift and break schedule
strategies?
•Where are the potential bottleneck areas?
•Which task assignment rules maximize utili-
•How many vehicles do you need to meet
activity requirements? zation?
•What is the best idle person deployment that
•What are the best scheduled maintenance/
minimizes response time?
recharging strategies?
•Which task assignment rules maximize vehi-
cle utilization? Industrial Vehicles
•What is the best idle vehicle deployment to
Industrial vehicles include all push or powered
minimize response time?
carts and vehicles that generally have free move-
•Is there any possibility of deadlocks?
ment. Utilize powered vehicles such as lift trucks
•Is there any possibility of collisions? for medium distance movement of batched parts
in a container or pallet. For short moves, use
Manual Material Handling Sys- manual or semi-powered carts. Single load trans-
porters are capable of moving only one load at a
tems time from one location to another. Such devices
Manual material handling systems (MMHS) are are fairly straightforward to model because they
probably the most common systems for handling, involve only a single resource and a single desti-
moving, storing, retrieving, and managing mate- nation for each move.
rials. ProModel easily allows you to model man- Multiple load transporters, on the other hand, can
ual labor. With shifts, breaks, and downtimes move more than one load at a time from one or
(illness, training, etc.), model people as more sources to one or more destinations. Defin-
resources. ing capacity and operation for multiple load
Whether carrying parts from one station to transporters can be extremely difficult since there
another or stocking shelves in a grocery store, may be special rules defining when to retrieve
people can be modeled as resources to do just additional loads and when to deliver the loads
about anything in ProModel. Such resources are already on board.
dynamic and move along assigned path networks.
Since people can think, you can model them very Modeling Industrial Vehicles
much like automated guided vehicles with on- Modeling an industrial vehicle involves modeling
board vehicle sensing. a resource that moves along a path network.
Paths are typically open aisles in which bi-direc-
MMHS Simulation Benefits tional movement is possible and passing is per-
Simulating your manual labor material handling mitted. You may also need to define deployment
system will help you address design and opera- strategies (work searches, idle vehicle parking,
tional issues, answering the following questions: etc.).
ProModel 283
User Guide

Since people generally operate industrial vehi- means that movement along one axis is complete
cles, they take breaks and may be available only before movement along the other axis begins.
during certain shifts. So in addition to modeling From a modeling standpoint, it is usually only
the vehicle, you may need to model an operator necessary to calculate and model the longest
that takes breaks and works on shifts or you may move time.
just want to assign shifts and breaks to the vehicle
In modeling AS/RS, the storage capacity is usu-
resource itself, in effect, modeling the vehicle
ally not a consideration and the actual inventory
and the operator as one resource.
of the system is not modeled. It would require
lots of overhead to model the complete inventory
Industrial Vehicle Simulation in a rack with 60,000 pallet locations. Since it is
Benefits primarily only the activity that is of interest in the
Simulating your industrial vehicles will help you simulation, actual inventory is ignored. In fact, it
address design and operational issues, answering is usually not even necessary to model specific
the following questions: stock keeping units (SKUs) being stored or
retrieved, but only distinguish between load type
•What is the required number of vehicles to
insofar as it affects routing and subsequent opera-
handle the required activity?
tions.
•What is the best deployment of vehicles to
maximize utilization?
•What is the best deployment of vehicles to Modeling an AS/RS as a Vertical
minimize response time? Bridge Crane
One way to model an AS/RS with greater accu-
racy is to use the bridge crane construct. An AS/
Automated Storage/Retrieval RS is really just a vertical crane with the same
Systems characteristics as a bridge crane.
An automated storage/retrieval system (AS/RS)
is “a combination of equipment and controls AS/RS Simulation Benefits
which handles, stores, and retrieves materials Simulating your AS/RS will help you address
with precision, accuracy, and speed under a design and operational issues, answering the fol-
defined degree of automation” (The Material lowing questions:
Handling Institute 1977). The goal of an AS/RS
•How many aisles do you need to handle the
is to provide random, high density storage with
required activity?
quick load access, all under computer control.
•What is the best sequence of stores and
retrievals to maximize throughput?
Modeling AS/RSs •What is the best stationing of empty S/R
At a simple level, an AS/RS move time may be machines to minimize response time?
modeled by taking a time from a probability dis-
tribution that approximates the time to store or
retrieve a load. More precise modeling incorpo-
rates the actual crane (horizontal) and lift (verti-
cal) speeds. Each movement usually has a
different speed and distance to travel which
284 Chapter 6:
Modeling Tips

Modeling Tips cess time at Machine_1. You can assign a value


to “Proc_time” in the Operation field or, as men-
The following section provides advanced Model- tioned above, in the Logic field in the Arrivals
ing Tips. module. The latter allows you to assign a value to
the attribute for each entity as they arrive in the
system.
Using Entity Attributes When using attributes instead of operation times,
Entity attributes are a powerful construct that ProModel substitutes the value assigned to the
enable a user to “tag” entities with characteristics attribute directly into the Operation field. So, if
to distinguish them from other entities of the “Proc_time=4.5” were your attribute assignment
same name. and the default units for time were in minutes, the
Entity attributes are place holders associated with resulting operation time would be 4.5 minutes.
individual entities which usually contain infor- You may also use attributes to determine the
mation about that entity. Attributes may contain routing of an entity. Note that the Routing Rule
integers or real numbers. Unlike variables, dialog offers a “User Condition” routing rule. In
attributes are not global. To define attributes, use this field, you can enter an expression that evalu-
the Attributes module found under the More Ele- ates the value of an attribute for the entity cur-
ments option of the Build menu. rently at that location. Hence, ProModel selects
One use of an entity attribute is to represent a the routing destination based on the attribute.
processing time. This is especially useful when
you base processing times at a location on some
characteristic other than entity type, such as size
or condition.
Suppose several types of tires enter a machining
center that places the steel belts on the tire. The
processing time of the tire increases as the diame-
ter of the tire increases. The following example
shows that for a tire entering a machining loca-
tion, the processing time uses the value of the
attribute “Proc_time.” You can make the attribute
assignment in the “Logic” field of the Arrivals
module.

In this example, ProModel attaches the attribute


“Proc_time” (short for process time) to the entity,
In this example, an entity attribute called
tire, which contains a value representing the pro-
“Inspected” has a value of zero or one. For unin-

You might also like