CHAPTER 3: WORK FLOW AND BATCH PROCESSING

IMSE-321: Work Design and Measurement Prof. Mustapha Nour El Fath

3.1 Sequential Operations and Work Flow

Unit operations: refers to the single tasks or operations performed at one location and independently of other operations. Sequential operations: refers to a work system in which multiple processing steps are accomplished in order to complete a work unit, and the processing steps are performed sequentially (rather than simultaneously).

Sequential operations

Unit operations

3.1 Sequential Operations & Work Flow

The work units may be materials, parts, products, or people In sequential operations, there are usually limitations on the order in which the operations can be performed (i.e., some operations must be completed before others can be started, for example, a hole must be drilled before it can be tapped to cut the threads), called precedence constraints

3.1 Sequential Operations & Work Flow

Example: passenger traveling in an airport

Operation 1: check in at the airline desk Operation 2: go through security Operation 3: find your travel gate Work unit that flows from one operation to another is the passenger

3.1 Sequential Operations and Work Flow

Associated with the sequential operations is work flow which refers to the physical movement of work units

3.1.1 Work Flow Patterns

3.1.1 Work Flow Patterns

Two basic types of work flow patterns can be distinguished: 1. Pure sequential pattern: all work units follow the same exact sequence of workstations and operations. There is no variation in the processing sequence. 2. Mixed sequential pattern: there are variations in the work flow for different work units. The different work units are processed through different stations.

3.1.1 Work Flow Patterns

Diagrams and charts used to show the flow of work units through a series of operations:
1. Network diagram 2. From-to chart

3.1.1 Work Flow Patterns

1. Network diagram: Consist of nodes representing operations, work elements, or other entities and arrows connecting the nodes indicating relationships among the nodes (i.e., work flow between nodes)
Arrow

Node = operations applied on the work unit Arrow = flow of work units

Node

3.1.1 Work Flow Patterns

Types of work unit move
In-sequence Bypassing Backflow Repeat operation

Move description
forward transport to operation immediately downstream forward transport to an operation beyond the neighboring station transport in a backward direction operation is repeated at the same workstation

3.1.1 Work Flow Patterns

2. From-To chart: is a table that can be used to indicate various quantitative relationships between operations or workstations in a multistation work system Possible variables in a From-To chart:

Quantities of work units moving between operations or workstations (Qij) Flow rates of materials moving between operations or workstations (Rfij) Distances between work stations (Lij)

i = from operation i j = to operation j

3.1.1 Work Flow Patterns

To operation j 1 1 2 From operation 3 i 4 5 Q11 Q21 Q31 2 Q12 Q22 Q32 3 Q13 Q23 Q33 4 Q14 Q24 Q34 5 Q15 Q25 Q35

Q41
Q51

Q42
Q52

Q43
Q53

Q44
Q54

Q45
Q55

Repeat moves In-sequence moves Bypassing moves Backflow moves

From-To Chart
To operation j 1 1 2 From operation i 3 4 5 2 40 10 30 20 25 50 3 4 15 5

Network Diagram
Network diagram showing same data as in previous From-To Chart

Problem 3.1: Network Diagram

Five jobs (J1, J2, J3, J4, and J5) are processed through a sequence of five operations (1, 2, 3, 4, and 5). Not all parts are processed in all operations. Job J1, which has daily quantities of 50 units, is processed through operations 1, 3, 5, and 1 in that order. Job J2, which has daily quantities of 70 units, is processed through operations 2, 4, and 5 in that order. Job J3, which has daily quantities of 25 units, is processed through operations 3, 2, and 4 in that order. Job J4, which has daily quantities of 10 units, is processed through operations 1, 2, 4, and 5 in that order. Finally, Job J5, which has daily quantities of 15 units, is processed through operations 3, 1, and 2 in that order. (a) Draw the network diagram and (b) prepare the From-To table for this work system.

Solution 3.1 : Network Diagram

Solution: (a) Network diagram
10+15 1 50 50 15 50 5 70+10 4 3 70+25+10 2 25

(b) From-to chart

From\To 1 2 3 4 5

2 25

3 50

105 15 25 50 80 50

3.1.2 Bottlenecks in Sequential Operations

3.1.2 Bottlenecks in Sequential Operations

In a work system consisting of a sequence of processing operations, the overall production rate of the system is limited by the slowest operation in the sequence which is called the bottleneck. Rps = Min {Rpi} for i = 1,2, .. , n

where Rps = overall production rate of the system, pc/hr Rpi = production rate of operation i, pc/hr n = the number of operations in the sequence

3.1.2 Bottlenecks in Sequential Operations

Ultimately, the slowest process limits the output of the other operations in the sequence. It may be technologically possible to run the other operations faster, at least those that are upstream from the bottleneck, but this would only cause an accumulation of parts in front of the bottleneck. In the long run, the upstream operations must produce at a rate that is no greater than the bottleneck operation. The upstream operations are said to be blocked Blocking means that the production rates of one or more upstream operations are limited by the rate of a downstream operation

3.1.2 Bottlenecks in Sequential Operations

The downstream operations can work no faster than the rate at which the bottleneck feeds work units to them. The operations downstream from the bottleneck are said to be starved for work Starving means that the production rates of one or more downstream operations are limited by the rate of an upstream operation (e.g., the bottleneck)

Example 3.1: Bottlenecks in Sequential Operations

There are 20 automatic turning machines in the lathe department. Batches of parts are machined in the department. Each batch consists of setup and run. Batch size = 100 parts. The standard time to set up a machine for each batch is 5.0 hours. Four setup workers perform the setups. They each work 40 hours per week. Once a machine is set up, it runs automatically, with no worker attention until the batch is completed. Cycle time to machine each part = 9.0 min; thus, it takes 15 hours of run time to produce a batch. Assume all machines are perfectly reliable. (a) What is the production output of the lathe department in 40 hours of operation per week?

Example 3.1: Bottlenecks in Sequential Operations

Solution: (a) Compare machine output capacity and setup output capacity. Machine capacity: Batch time TB = 5.0 + 100(9/60) = 20 hr/batch In a 40 hr week, each machine produces 2 batches = 2(100) = 200 pc/wk For 20 machines, Q = 20(200) = 4000 pc/wk Setup worker capacity: With Tsu = 5.0 hr, in 40 hr, each setup worker can set up 40/5 = 8 setups/wk. With 4 setup workers, number of setups = 4(8) = 32 setups = 32 batches Output is limited by setup worker capacity to Q = 32(100) = 3200 pc/wk

Example 3.2: Bottlenecks in Sequential Operations

A factory produces one product. One unit of raw material is required for each unit of product. Two processes are required to produce the product, process 1, which feeds into process 2. A total of five identical machines are available in the plant that can be set up to perform either process. Once set up, each machine will be dedicated to perform that process. For each machine that is set up for process 1, production rate = 12 units per hour. For each machine that is set up for process 2, production rate = 18 units per hour. Both processes produce 100% good units (fraction defect rate = 0). A work-in-process buffer is provided between the two processes to avoid starving and blocking of machines. The factory operates 40 hours per week. (a) In order to maximize factory production, how many machines should be set up for process 1, and how many machines should be set up for process 2? (b) What is the factorys maximum possible weekly production rate of good product units?

Example 3.2: Bottlenecks in Sequential Operations

Solution: (a) Rp1 = 12 pc/hr, Rp2 = 18 pc/hr n1 = (18/12)n2 = 1.5 n2 n1 + n2 = 1.5n2 + n2 = 2.5n2 = 5 machines total n2 = 2 machines, n1 = 5 2 = 3 machines (b) For the two operation sequence, Rp = 2(18 pc/hr)(40 hr/wk) = 1440 pc/wk Check: Rp = 3(12 pc/hr)(40 hr/wk) = 1440 pc/wk

3.2 Batch Processing

Batch processing: it consists of the processing of work units (e.g., materials, products, information, or people) in finite quantities or amounts Work units are often processed in batches Examples:
Passengers who travel by airplane are transported in batches Freight is moved in batch loads by truck or railway train

3.2 Batch Processing

Types of Batch Processing
Sequential batch processing: Members of the batch are processed one after the other (e.g., grading of student papers) Simultaneous batch processing: Members of the batch are processed all at the same time (e.g., passenger air travel)

3.2.1 The Pros and Cons of Batch Processing

3.2.1 The Pros and Cons of Batch Processing

Disadvantages of batch processing:
Time lost in equipment setup (i.e., lost production time)
Example: time lost in changing the machine tool for the next part style

Work-in-process (WIP): the accumulation of large quantities or amounts of work units in the sequential processing system
Example: the accumulation of work units in front of a machine

3.2.1 The Pros and Cons of Batch Processing

Despite the disadvantages cited above, batch production is nevertheless widely used for the following reasons:
Work unit differences: there are differences in work units between batches, and it is necessary to make changes in the methods, tooling, and equipment to accommodate the differences Equipment limitations: the size capacity of the equipment restricts the amount of material or quantity of work units that can be processed at one time (e.g., the equipment capacity imposes an upper limit on the batch size)

3.2.1 The Pros and Cons of Batch Processing

Alternating cycles of setup and production run experienced by a work system engaged in batch production

3.2.1 The Pros and Cons of Batch Processing

While the work system is producing, its production rate is greater than the demand rate for the current product type This means that the units in a batch of items must be held in inventory for extended periods of time, while demand gradually reduces the stock level down to the point at which another production run will be made This is the typical make-to-stock situation, in which items are manufactured to replenish inventory that has been gradually depleted by demand

3.2.1 The Pros and Cons of Batch Processing

An important question arises in make-to-stock situations and in batch production: How many units should be produced in a given batch? The answer involves achieving a balance between inventory costs and setup costs

3.2.1 The Pros and Cons of Batch Processing

3.2.2 Economic Order Quantity Model

3.2.2 Economic Order Quantity Model

The number of units produced in a given batch must minimize the inventory (i.e., holding) costs and setup costs Inventory or holding costs emerge from the money the company must pay for the space occupied by the inventory Setup costs emerge from the cost of the labor performing the setup changes

3.2.2 Economic Order Quantity Model

The total annual inventory cost is

ChQ Csu Da TIC 2 Q

where Q = order quantity, pc/order TIC = total annual inventory cost in $/yr Da = annual demand for the item, pc/yr Csu = setup cost, $/setup Ch = inventory holding cost, $/pc/yr Note that Da/Q represents the number of orders or batches produced per year; it therefore gives the number of setups per year

3.2.2 Economic Order Quantity Model

The total annual cost (including the actual annual cost of part production) is

ChQ Csu Da TC Da C pc 2 Q
where Q = order quantity, pc/order TC = total annual cost in $/yr Da = annual demand for the item, pc/yr Csu = setup cost, $/setup Ch = inventory holding cost, $/pc/yr Cpc = unit cost of the item, $/pc Note that Da/Q represents the number of orders or batches produced per year; it therefore gives the number of setups per year

3.2.2 Economic Order Quantity Model

The holding cost (Ch) = h Cpc where Ch = holding cost, $/pc/yr h = holding cost rate, rate/yr = i(interest rate) + s(storage rate) Cpc = unit cost of the item, $/pc The setup cost (Csu) = Tsu Cdt where Csu = setup cost, $/setup Tsu = setup time between batches, hr/setup Cdt = cost rate of machine downtime during the changeover, $/hr

3.2.2 Economic Order Quantity Model

The batch size that minimizes the sum of carrying costs and setup costs is:
Q = EOQ = 2 Da Csu Ch

where Q = order quantity, pc/order EOQ = economic order quantity (number of parts to be produced per batch), pc/batch Da = annual demand for the item, pc/yr Csu = setup cost, $/setup Ch = inventory holding cost, $/pc/yr

3.2.2 Economic Order Quantity Model

Example 3.1: Economic Order Quantity Formula The annual demand for a certain item made-tostock is 15,000 pc/yr. One unit of the item costs $20.00 and the holding cost rate is 18%/yr. Setup time to produce a batch is 5 hr. The cost of equipment downtime plus labor is $150/hr. Determine the economic order quantity (EOQ) and the total inventory cost for this case.

3.2.2 Economic Order Quantity Model

Solution: Setup cost (Csu) = Tsu Cdt = 5 $150 = $750. Holding cost (Ch) = h Cpc = 0.18 $20.00 = $3.60. Using these values and the annual demand rate in the EOQ formula, we have

2 Da Csu 215000750 EOQ 2500 units Ch 3.60 Total inventory cost is given by the TIC equation

ChQ Csu Da 3.602500 75015000 TIC $9000 2 Q 2 2500

3.2.2 Economic Order Quantity Model

Solution: Including the actual production costs in the annual total ChQ Csu Da TC Da C pc 2 Q 15,000(20) 9000
$309,000

Example 3.3 EOQ Model

Annual demand for a certain part = 10,000 units. At present the setup time on the machine tool that makes this part = 5.0 hr. Cost of downtime on this machine = $200/hr. Annual holding cost per part = $1.50. Determine (a) EOQ and (b) Total inventory costs for this data. Also, determine, if the changeover time could be reduced to six minutes. c) EOQ and d) Total inventory costs

Example 3.3 EOQ Model

3.3 Defects in Sequential Operations and Batch Processing

In a sequence of operations, defective units maybe produced in any or all of the operations The defect rate must be considered in determining the quantity of good units produced

3.3 Defects in Sequential Operations and Batch Processing

Q = Qo(1-q) D = Qo q
Incoming units Good products and defects

where Qo = starting quantity Q = quantity of good units produced q = fraction defect rate D = number of defects produced

3.3 Defects in Sequential Operations and Batch Processing

Qo = starting quantity Qf = quantity of good units produced qi = fraction defect rate in operation i Df = number of defects produced

Sequence of operations

3.3 Defects in Sequential Operations and Batch Processing

Good units produced after n operations: Qf = Qo(1 q1)(1 q2) . . (1 qn) Resulting defects: Df = Qo [1-(1 q1)(1 q2) . . (1 qn)]

Yield for the sequence (the ratio of good units produced to starting units): Y = Qf / Qo = (1 q1)(1 q2) . . (1 qn)

3.3 Defects in Sequential Operations and Batch Processing

Example 3.2: The Compounding Effect of Defect Rate in Sequential Operations A starting batch of 1000 work units is processed through 10 operations, each of which has a fraction defect rate of 5%. Determine (a) how many good parts and defects are produced by the first operation, (b) how many good parts and defects are in the final batch, and (c) the yield of the first operation and the yield of the operation sequence.

3.3 Defects in Sequential Operations and Batch Processing

Solution: (a) For the first operation, Qf1 = Qo (1-q) = 1000(1-0.05) = 950 good units Df1 = Qoq = 1000(0.05) = 50 defects

3.3 Defects in Sequential Operations and Batch Processing

Solution: (b) For the 10 sequential operations, Qf = Qo(1 q1)(1 q2) . . (1 q10) = Qo(1 q)10 = 1000(1-0.05)10 = 599 good units

Df = Qo [1-(1 q1)(1 q2) . . (1 q10)] = Qo [1-(1 q)10] = 1000 [1-(1-0.05)10] = 401 defects

3.3 Defects in Sequential Operations and Batch Processing

Solution: (c) The yield of the first process is Y1 = Qf1 / Qo = 950 / 1000 = 95% The yield of the process sequence is Y = (1 q)10 = (1-0.05)10 = 0.5987 = 59.87%

Problem 3.20
Three sequential operations are required for a certain automotive component. Operation 1 has a defect rate = 4%. Operation 2 has a defect rate = 5%. Operation 3 has a defect rate = 10%. Operations 2 and 3 can be performed on units that are already defective. If 25,000 starting parts are processed through the sequence, (a)how many units are expected to be defect-free, (b)how many units are expected to have exactly one defect, and (c)how many units are expected to have all three defects?

Solution 3.20
a) Number of defect-free units = 25,000(1 0.04)(1 0.05)(1 0.10) = 25,000(0.96)(0.95)(0.90) = 25,000(0.8208) = 20,520 pc (b) Number of units with one defect
D1 = 25,000(0.04)(1 0.05)(1 0.10) = 855 pc D2 = 25,000(1 0.04)(0.05)(1 0.10) = 1080 pc D3 = 25,000(1 0.04)(1 0.05)(0.10) = 2280 pc D1 + D2 + D3 = 4215 pc

c) Number of units with all three defects D123 = 25,000(0.04)(0.05)(0.10) = 5 pc

Problem 3.24
Two sheet metal parts, A and B, are produced separately, each requiring two press-working operations. Part A is routed through operations 1 and 2, and part B is routed through operations 3 and 4. The two parts are then joined in a welding step (operation 5), and the assembly is routed to an electroplating operation (operation 6). The six operations have the following fraction defect rates: q1 = 0.05, q2 = 0.15, q3 = 0.10, q4 = 0.20, q5 = 0.13, q6 = 0.08.

If the desired final quantity of assemblies is 100,000 units, how many starting units of parts A and B will be required? There is no inspection or separation of defective units until after the final process, so defective units and good units are processed together through all production processes.

3.24 Solution
Required output of Q6 = 100,000 pc
Qo (A) 1 q = 0.05 2 q = 0.15 5 q = 0.13 Qo (B) 3 q = 0.10 4 q = 0.20 6 q = 0.08 Q6 = 100,000

Q6 = Q5(1 0.08), Q5 = 100,000/(0.92) = 108,696 pc With q5 = 0.13, Q2 = Q4 = 108,696/(1 0.13) = 124,938 pc Starting quantity Qo(A) = 124,938/{(1-0.15)(1-0.05)} =154,722 Starting quantity Qo(B) = 124,938/{(1-0.20)(1-0.10)} = 173,525

3.4 Work Cells and Worker Teams

A work cell is a group of workstations dedicated to the processing of a range of work units within a given type The work units are parts and the range of parts are called part family The members of the part family possess similarities that permit them to be processed by the work cell Work cells and part families are associated with group technology

Group Technology
An approach to manufacturing in which similar parts are identified and grouped together to take advantage of their similarities in design and production Work units are processed individually and continuously, without the need for time-consuming changeovers between part types Requires: parts with a certain degree of similarity and adaptable/flexible equipments/workers Avoids disadvantages of batch processing (lead times and WIP)

3.4 Work Cells and Worker Teams

Work Cells Advantages:
Reduced material handling: move distances and move times within a cell are negligible because of the next required machines close proximity Reduced setup time: due to the similarities in work units, few changes in the machines tools will reduce the setup time Reduced work-in-process: the smooth and continuous process of work units without excessive setup time will reduce the work-inprocess

3.4.1 Work Cell Layouts and Material Handling

3.4.1 Work Cell Layouts and Material Handling Material Handling Methods
The number of workstations in a cell can range from two to about a dozen. There are no hard limits on the upper end of the range If the number of workstations is very large (e.g., several dozen to several hundred), then the work is more likely to be organized as a manual assembly line

3.4.1 Work Cell Layouts and Material Handling Material Handling Methods
There are two types of material handling methods in work cells:
1. Manual 2. Mechanized

3.4.1 Work Cell Layouts and Material Handling Material Handling Methods
1. Manual: consists of the workers in a cell moving the work units between stations

3.4.1 Work Cell Layouts and Material Handling Material Handling Methods
2. Mechanized: work units movement is achieved by means of powered conveyor

3.4.1 Work Cell Layouts and Material Handling

In terms of the cell layout, work cells can be categorized as:
1. 2. 3. 4. In-Line U-Shaped Loop Rectangular

3.4.1 Work Cell Layouts and Material Handling In Line Work Cell

In-line work cell layout is often the best flow pattern for long, narrow buildings In-line work cell layout is used when the shipping and receiving sections are on the opposite sides

3.4.1 Work Cell Layouts and Material Handling U-shaped Work Cell

The U-shaped work cell layout can be placed on a small area (i.e., save space) The U-shaped work cell layout requires less walking than other layouts

3.4.1 Work Cell Layouts and Material Handling Loop Work Cell

3.4.1 Work Cell Layouts and Material Handling Rectangle Work Cell

3.4.2 Determining the Operation Sequence in a Work Cell

3.4.2 Determining the Operation Sequence in a Work Cell

What is the most appropriate sequence of workstations in a work cell? Hollier technique is one of the available techniques for determining the most appropriate sequence of workstations in a work cell The Hollier technique objective is to arrange the stations in an order that maximizes the proportion of forward moves (in sequence and bypassing moves) within the cell

3.4.2 Determining the Operation Sequence in a Work Cell

Hollier method can be outlined as follows:
Develop the From-To chart from part routing data Determine the From-To ratio for each workstation Arrange workstations in order of decreasing From-To ratio Note: in case of a tie, the workstation with the higher From value is placed ahead of the station with a lower value

3.4.2 Determining the Operation Sequence in a Work Cell

Example 3.3: Work Cell Station Sequence Four workstations, 1, 2, 3, and 4, have been assigned to a work cell. An analysis of 50 parts processed in these stations has been summarized in the following From-To chart. Additional information is that 50 parts enter the cell at station 3, 20 parts leave after processing at station 1, and 30 parts leave after station 4. Determine a logical workstation arrangement using the Hollier algorithm.

3.4.2 Determining the Operation Sequence in a Work Cell

From-to Chart To 1 2 3 1 0 30 10 2 5 0 40 3 0 0 0 4 25 15 0

From

10

3.4.2 Determining the Operation Sequence in a Work Cell

From-to Sums and Ratios To 1 From 2 3 4 To sums 1 0 30 10 10 50 2 5 0 40 0 45 3 0 0 0 0 0 4 25 15 0 0 40 From sums 30 45 50 10 135 From-to ratios 0.60 1.0 0.25

Solution: From the above table, arranging the stations in order of descending From-To ratio, the cell is sequenced as follows: 3 2 1 4

3.4.2 Determining the Operation Sequence in a Work Cell

10 50 40 30 15

In

25

30

Out

10

20

Out

3.4.2 Determining the Operation Sequence in a Work Cell

Several performance measures can be used to compare alternative solutions to the machine sequencing problem. The measures are based on the following moves:
Percentage of in-sequence moves, computed by adding all of the values representing in-sequence moves and dividing by the total number of moves Percentage of bypassing moves, found by adding all of the values representing bypassing moves and dividing by the total number of moves Percentage of backflow moves, determined by summing all of the values representing backflow moves and dividing by the total number of moves Percentage of repeated operations, which is the sum of all repeated operations divided by the total number of moves

3.4.2 Determining the Operation Sequence in a Work Cell

It is desirable for the layout arrangement to have high proportions of in-sequence and bypassing moves since these both represent forward work flow (in-sequence moves are more desirable than bypassing moves) The layout should minimize the percentage of backflow moves

3.4.2 Determining the Operation Sequence in a Work Cell

Example 3.4: Performance Measures for Example 3.3 Compute (a) the percentage of in-sequence moves, (b) the percentage of bypassing moves, and (c) the percentage of backflow moves for the solution in example 3.3

3.4.2 Determining the Operation Sequence in a Work Cell

Solution: From the network diagram in example 3.3, the total number of moves is 135 (a) the number of in-sequence moves = 40+30+25 = 95; percentage of in-sequence moves = 95/135 = 0.704 = 70.4% (b) the number of bypassing moves = 10+15 = 25; percentage of bypassing moves = 25/135 = 0.185 = 18.5% (c) the number of backflow moves = 5+10 = 15; percentage of backflow moves = 15/135 = 0.111 = 11.1%

3.4.3 Worker Teams

3.4.3 Worker Teams

A worker team is a group of employees who work together to achieve common objectives In the case of a work cell, the common objectives of the team are to:

1. Meet the production or service schedule 2. Achieve high quality in the goods or services provided by the cell 3. Make the operation of the cell as efficient as possible

These objectives are achieved by means of teamwork, in which the collective skills and efforts of the team members exceed the sum of their individual skills and efforts

Teamwork the collective skills and efforts of the team members exceed the sum of their individual skills and efforts

Cross-training workers become trained in more than one job in the cell Allows for job rotations to increase work variety and job satisfaction Mitigate problems of absences