Zagazig University, Faculty of Engineering,

Mechatronics Program

“Manufacturing Management”
Case study

By: Hazem tawfeek yassen tawfeek

Ahmed tarek hamed ahmed
Youssef tarek abdellatif
EZZ Eldin mohammed

Level: 300
Under supervision of : Dr. Mansour Abogamila
 Operations Scheduling Real Case Study

Abstract

In this paper, we analyzed the sequencing situations on two

machines where the machine setup time is not independent of
processing order.
A real case study of Badr City Industrial Complex is taken as a
model. Data is collected and analyzed using MS-Excel by
different methods. The problem formulation has been
presented. Multiple solutions were obtained by applying
sequencing methods. The comparison of different solutions is
done to choose the optimal solution. The time is reduced by
23% to perform the group of jobs and the setup time is reduced
30.5% as well as the mean flow time is reduced by 30.5%.
1. Introduction 3. Scheduling criteria,
4. Parameter variability,
Flow shop scheduling is one of the 5. Scheduling environment.
most important problems in the area of In the literature, there are many
production management. It can be papers published in which the sequencing
briefly described as follows: There are issue is tackled and investigated.
a set of m machines (processors) and a Aggarwal (1975) presented a scheduling
set of n jobs. Each job comprises a set algorithm to solve
of m operations which must be done on flowshop problems with a
different machines. All jobs have the common job sequence on all machines.
same processing operation order when This algorithm used makespan as its
passing through the machines. There criterion and offered up to 1% average
are no precedence constraints among improvement in reducing the makespan
operations of different jobs. Operations of nearly 50% of the problem sets over
cannot be interrupted and each the results of the existing algorithms.
machine can process only one Caffrey and Hitchings (1995)
operation at a time. The problem is to considered scheduling of five jobs
find the job sequences on the machines through a flow shop with five machines.
which minimize the make span, i.e. the They obtained the distribution of make
maximum of the completion times of spans and the distribution of the optimal
all operations. As the objective make spans by complete enumeration of
function, mean flow time, completion all the schedules. Torres and Centeno
time variance and total tardiness can (2008) considered a permutation flow
also be used. According to Weng and shop problem with secondary resources
Haiying (2006) , the flow shop with the objective of minimizing the
scheduling problem is usually solved number of tardy jobs. He presented a
by approximation or heuristic methods. lower bound for the permutation
These methods ranged from Gantt flowshop problem and evaluates its
charts and the assighment methods of performance against the optimal solution
scheduling to a series of priority rules, for small, medium, and large instances.
the critical – ratio rule, Johnson’s rule Weng and Haiying (2006) presented a
for sequencing, and finite capacity priority rule for dynamic job shop
scheduling. (Heizer and scheduling that minimizes mean job
Render(2006)). tardiness.
Scheduling has been defined as Chan et. al. (2005) developed an
'the art of assigning resources to tasks assignment and scheduling model to
in order to insure the termination of study the impact of machining flexibility
these tasks in a reasonable amount of on production issues such as job lateness
time'. According to Voss et. al. (2002), and machine utilization and suggested an
the general problem is to find a improvement of overall production
sequence, in which the jobs pass performance if the equilibrium state can
between the resources, which is a be quantified between scheduling
feasible schedule, and optimal with performance and capital investment. Also
respect to some performance criterion. Chan et. al. (2006) solved a resource-
Blazewicz (2005) introduced a constrained operations–machines
functional classification scheme for assignment problem and flexible job-
scheduling problems. This scheme shop scheduling problem iteratively.
categorizes problems using the Konstantin et. al. (2005) focused
following dimensions: on a dynamic generalization of the
1. Requirement generation,
assignment problem where each task
2. Processing complexity,
consists of a number of units to be
performed by an agent or by a limited
number of agents at a time. important pieces of information. Job time
Also, Gupta et. al. (2004) usually includes setup and processing
considered a variant classical problem times. Jobs that require similar setups can
of minimizing makespan in a two- lead to reduced setup time if sequencing
machine flow shop. In this variant, rules are taken into account. Also, it
each job has three operations, where should be noted that the priority rules can
the first operation must be performed be classified as either local or global.
on the first machine, the second Local rules are taken into
operation can be performed on either account for information pertaining only
machine but cannot be preempted, and to a single machine while global rules are
the third operation must be performed taken into account for information
on the second machine. They showed pertaining to more than one machine.
that a schedule in which the Moreover, a number of following
operations are sequenced assumptions are applied when we use
arbitrarily, but without inserted priority rules:
machine idle time, has a worst-case 1. The set of jobs is known, no jobs
performance ratio. arrive after processing begins,
Recently, Agarwal (2006) and no jobs are cancelled.
proposed a meta-heuristic approaches 2. Setup time is independent of
for the two-machine flow-shop processing sequence.
problem with weighted late work 3. Setup time is deterministic.
criterion and common due date. Also, 4. Processing times are
Petrovic and Song (2003) introduced a deterministic.
new approach to two-machine flow The effectiveness of any given
shop problem with uncertain sequence is frequently judged in terms of
processing time. In their paper, flow one or more performance measures. The
shop problem with uncertain most frequently used performance
processing time was represented with measures are, job flow time, job lateness,
fuzzy number. Especially, the scheme makespan and average number of jobs.
used in McCahon and Lee’s algorithm Johnson described a heuristic
for ranking fuzzy processing times was method that can be used for the case
modified to calculate better minimum where a set of jobs is to be processed
makespan. through two machines. In this technique,
In this paper, we analyzed the managers can use to minimize
sequencing situations under two makespan for a group of jobs to be
machines where the machine setup processed on two machines or at two
time is dependent taken into account successive work centers (sometimes
for the real industrial company. referred to as a two-machine flow shop).
1.1 Priority Rules and Assumptions For the technique to work, several
In the literature, a number of conditions must be satisfied:
priority rules are simple heuristics used 1. Job time including setup and
to select the order in which jobs will be processing must be known and
processed. Such rules are given in the constant for each job at each
Table 1. work centre.
The rules generally rest on the 2. Job times must be independent of
assumptions that job setup time and the job sequence.
cost are independent of processing 3. All jobs must follow the same
sequence. By using these rules, job two-step work sequence.
processing time and due dates are 4. Job priorities cannot be used.
 5. All units in a job must be
completed at the first work then reset. Operators have to shift jobs,
center before the job moves on often moving from one location to
to the second work center. another. The learning curve comes into
1.2 The Theory of Changeovers play every time operators bring a new
The preceding discussion order online which is also a part of the
assumed that all the rules are used changeover process.
where the setup time is independent of Finally, we should organize that
processing order, but in many setting up times and differ according to
instances that assumption is not true. requirement of each job and whether a
Consequently, a manager may want to system is designed for high volume, or
schedule jobs at a machine taking into low volume. Some jobs may need to
account those dependencies. The goal change a specific tools and equipment on
is to minimize total setup time or a particular machine while other jobs
changeovers. may need to replace some devices to
Changeovers are the steps that operate. Consequently, the problem of
need to be taken to prepare equipment sequencing will be more complicated by
and worker to do a new job. The term the variable number of each job in terms
setup is usually applied to the start up of processing time and coordination of
of a new job, which means the cleanup setup times.
from the old one. For process analysis, 2. A Model
it is best to separate these two steps
and deal with the setup and put away As shown in previous section of
(cleanup) as two activities. this paper, sequencing can be difficult for
The critical importance to job a number of reasons. One is that in
shop processes is the changeover cost reality, an operation must deal with
and time. These are often treated as variability in setup time, processing
being synonymous so that mention of times, changes in the set of jobs.
cost implies and vice-versa. Thus, the For a description of a heuristic that can
generic term setup is often used to be used for the case where a set of jobs is to be
include time and cost everything that processed through one machine given the setup
has to be done to change the process time. Consider the following table which shows
from one product to another. workstation machine setup times based on job
Setup costs are often processing order:
proportional to setup times. However,
the relationship breaks down when a
lot of technology is devoted to allow Resulting
very rapid setups. Then there is a cost following
of not using this technology for the Job setup
purposes for which it was intended, time (min)
namely, short runs of many designs is
that fit within the family of parts Setup
(technology group) that can be made time J1 J2 J3
on this equipment. Further, when the If the (minutes
setup can take place off-line, it can processing )
take longer and still cost much less job is J1 6 - 12 4
than when it must interrupt the J2 4 2 - 8
production process. Note if job J1 J3
is followed4 by 10
J2, the 6
setup-
Machines have to undergo time for J2 will be 12 minutes.
cleanup prior to job processing and Furthermore, if job J1 is completed first,
followed by job J2, job J3 will then
follow job J2 and have a setup time of PTij : Processing Time of Job i on
8 minutes. Then if job J1 is done first, Machine j
its setup time will be 6 minutes.
N : Number of jobs to be completed
The simplest way to determine M : Number of Machines in the
which sequence will result in the workshop
lowest total setup time is to list each TI ij : Starting Time for Job i on
possible sequence and determine its
Machine j (Time in)
total setup time. In general, the number
of different alternatives is equal to n!. STij : Setup time for Machine j where Job
Here, n is equal to 3, so there are six i is performed on it
alternatives and their total setup time is TSTij : Total Setup time for Machine j
as follows:
where all jobs are performed on it
(i=1,2,...,n)
Sequencing Total
Setup TTSTij : Total Setup time for all
time machines where all jobs are performed
(minutes on them
) (i=1,2,...,n and
J1-J2-J3 26 j=1,2,...,m)
J1-J3-J2 16
n
J2-J1-J3 10
J2-J3-J1 22
J3-J1-J2 26
J3-J2-J1 12

This procedure is relatively

simple to do manually when the j=1,2,...,m
TSTij
number of jobs is two or three.
However, as the number of jobs  STij
i 1
…..(1)
increases, the list of alternatives m n

quickly becomes larger, since if the

TTSTij  STij
number of machines is more than one. j 1 i 1
Thus, sequencing will be difficult for …..(2)
this reason. Z: Summation of total setting up times
In this study, we will analyze for all the machines for performing n
sequencing situations under two jobs.
machines where the setup time would The objective function Z will give the
be considered as a factor. total setting up time for all machines to
3. Problem Formulation complete all the jobs.
To form the general model Now we can form the problem of
subject to the setup time, the procedure sequencing if we consider two machines
for two machines is considered for which will be fit with our case study
setup time as a main factor influencing taken from an industry.
the sequence of a set of jobs. The Thus, the cells in the setup time
following notations are used to design matrix will differ accordingly based on
the model: job processing order, which job follows
and which job immediately predecessor.
Our objective is to find the minimum
total setup time on all machines for all By scheduling the jobs in the order of
the jobs, which can be represented as Machine 1 (Vertical Cutting) and
follows: Machine 2 (Circular Cutting), we
calculated the total setup cost for all jobs.
The complete summary of the calculation
for the existing system is shown in the
m n Table 6.
TTSTij  STij .(3)
j 1 i
1 In the Table 6, the following notations
are used:
4. Implementation
Notation Meaning
To implement the formula and M1 and Vertical Cutting
to achieve the above objective, a real M2 Machine and Circular
case study has been taken from Cutting Machine
industry. A group of ten jobs are to be T(M1) Time of processing for
processed through two machines flow and a job for Machine 1 and
shop. The first operation involves T(M2) Machine 2 respectively
Vertical Cutting and the second Ti Time in of a job on a
operation involves Circular Cutting. In machine
the existing system, the set of jobs are To Time out of a job on a
processed in the same order in which machine
they reached the department. To ST1 Setup time required for
implement the model, the jobs are Machine 1 to process a
labeled in serial number as they arrived job
to the flow shop. The respective ST2 Setup time required for
processing time for each job is given in Machine 2 to process a
the Table 3. Jobs are listed in order of job
arrival and the processing time is in TST Total Setup time
hours. required for two
The Table 4 and Table 5 machines to process a
contain order dependent setup times job
for all jobs on Vertical Cutting
Machine and Circular Cutting Machine
respectively. The data shows the setup It is clear from the Table 6 that all the
time for each machine in the flow ten jobs must be processed by Machine 1 and
shop. These times are in minutes. Note Machine 2 in 72 and 68 hours respectively but
that the time for undergoing cleanup due to idle of the machines, all the jobs will be
prior job processing and then be reset finished in 72 and 94 hours.
is included.

5- Data Analysis Total setup time for Machine 1

and Machine 2 for all jobs will be 118
5.1. Existing System minutes and 167 minutes respectively. So
the total setup time for both machines
In the existing system, the required to process all the jobs in the
production manager orders the jobs existing sequence will be 285 minutes. In
arbitrary and mainly using First Come this sequence, the mean flow time is 65.3
First Served (FCFS) rule, i.e. the jobs (653/10) hours.
are going to machine for processing in
the order in which they are arriving.
5.2. Searching The Optimum
Sequence completion time. But if production
By applying the efficient manager considers the setup time, it
algorithm suggested by Johnson (1954) varies as well as the mean flow time as
for solving two machine problems, shown in the tables.
multiple optimum sequences are Each sequence has different
generated. Each sequence is like an order of jobs which can be represented in
array of size 10 cells starting from left a Gantt chart. The Gantt chart in Figure 1
to right. These four sequences obtained is for sequence S3. This Gantt chart is
are labeled as S1, S2, S3, and S4. The prepared by using POM software which
order of the jobs in each sequence is is very useful to demonstrate how the
shown inside the chart. jobs are carried out. The time scale on
These four sequences are each machine is shown as starting and
generated from approximately ending time. Inside the chart, jobs are
3628800 permutation sequences. Here, written in order for sequence S3. Each
we have to note that in theory, a job has to wait for Machine 2 until it is
solution by enumeration is always free. For example, Job 8 (J8) is finished
possible, but in practice, the from Machine 1 after 20 hours and it will
computation of effectiveness for a not go directly to Machine 2 until it is
given sequence can be quite involved free. So, Machine 2 will be free after 31
and the number of cases for prohibitive hours and J8 has to wait until 31 hours
even for moderate number of machines for Machine 2. Opposite to this situation,
. So, by using Johnson's algorithm, sometimes the machine has to wait for
only sequences which are optimal in jobs. As happened for Machine 2, it will
terms of the total time (completion finish from Job 7 (J7) at 65 hours while it
time) to process all the group of jobs will not start to perform Job 1 (J1) until
on both machines is minimum among J1 is finished from Machine 1. That
all possible sequences generated. means, Machine 2 has to wait 3 hours.
So far we computed the This time is known as idle time and it is
relative result for the existing order. marked as shadow.
Now we extend the analysis and carry 6. Comparison Of The Study
on to find the optimal sequence in
terms of minimum total setup time The result for existing and proposed
(Min Z). So we consider each optimum system is summarized in the Table 12.
sequence which has been obtained. We It can be seen from the Table 12 that
calculated the compilation time to  the time taken by Machine 1
perform all the jobs and computed the (Vertical Cutting) is 72 for all the
mean flow time as well as setup time sequences as well as by existing
using MS-Excel . sequence, while time taken by
The calculation is based on Machine 2 (Circular Cutting) is
Table 4 and Table 5. The results are 94 for existing which is reduced
presented in the following tables i.e. to 23% for each proposed
Table 8, Table 9, Table 10 and Table sequence,
11 for the sequences S1, S2, S3 and S4
respectively.  the setup time is drastically
It is obvious from each solution reduced by approximately 44%
that the total time to complete all the on Machine 1 and 21% on
jobs is 73 hours for all the solutions. Machine 2 for the proposed
Hence each solution is an optimum solution (S3) as compared to the
solution in case of minimum existing one.
 Moreover, it is obvious that the
third sequence generates the optimal processors and the processing
solution according to the different time does not depend on the job.
factors. It gives the minimum time to
finish all the jobs and contains lowest 2. Somewhat more interesting
setup time for all the jobs on both results exist for scheduling jobs
machines (Min Z). The significant when considering scheduling as a
reduction occurs in mean flow time by dynamic problem; one must
almost 46.5%. Although, the study determine the pattern of arrivals
generated four multi-solutions by MS- to the system. It is common for
Excel but the third sequence (S3) gives jobs to arrive according to some
the best optimal solution in terms of random process and queue up for
setup time and completion time. service. Queuing theory and
7. Conclusions simulation may be useful as a
tool for dealing with randomness
In this paper, a model for of this type.
calculating the setup time is suggested. 3. There are a number of actions
A case study of real life is analyzed by that managers can be considered
using MS-Excel by different methods. to minimize sequencing problem
A significant result is generated which such as focusing on bottleneck
gave four solutions equally in operations, one can try to
completion and mean flow time but increase the capacity of the
they are differ in terms of setup time. operations if that is possible or
A unique solution (sequence S3) feasible, schedule the bottleneck
comes with the amazing result with operation first, and then schedule
minimum setup time among all four the non-bottleneck operations
sequences. around the bottleneck operations.
The major findings of this Thus, there is a need to develop a
study are generating different solutions method for identifying the
with equally lowest compilation time optimal schedule.
to finish all the group of jobs. The 4. Finally, there is a need to study
comparison of different solution is on an action that manager can
done to choose the optimized solution. consider minimizing scheduling
Particularly, for the sequence S3, the problems that is, considering the
time is reduced by 23% and the setup lot splitting for large jobs. This
time is reduced by 30.5% (87/285) for probably works best when there
both machines as well as the mean are relating large differences in
flow time is reduced by 30.5%. job timings.
8. Recommendations For References
Further Work
Agarwal, A., Colak, S., and Eryarsoy, E.,
"Improvement heuristic for the flow-shop scheduling
Based on the empirical findings of this
problem: An adaptive-learning approach", European
study, following are several points that
can be tackled later since we have Journal
not of Operational Research, 169(3) (2006) 801-
discussed in this paper: 815.
1. A class of problems that we
did not discuss but for which
there are several interesting
results, are problems in which
jobs are to be processed
through m non-identical
[2] Aggarwal, S.C. and Stafford, E. "A
heuristic algorithm for the [8] Heizer, J. and Render, B.,” Principles
flowshop problem with a common of operations Management” , Pearson
job sequence on all machines, education. Inc., New Jersey, 6 th. Ed.
Decision Sciences", 6(2) (1975) 2006.
237-251. [9] Konstantin, K., Eugene, K.
[3] Blazewicz, J., Pesch, E., and Toshihide, I., Dynamic Generalized
Sternaand, M. and Werner, F., "A Assignment Problems with Stochastic
comparison of solution procedures Demands and Multiple Agent--Task
for two-machine flow shop Relationships, Journal of Global
scheduling with late work Optimization, 31(1) (2005) 17-43.
criterion", Computers and [10] Petrovic, S. and Song, X., "A
Industrial Engineering, 49(4) new approach to two-machine flow
(2005) 611-624. shop problem with uncertain
[4] Caffrey, J. and Hitchings, G., processing time", 4th International
"Makespan distributions in flow Symposium on Uncertainty
shop scheduling", International Modeling and Analysis, (2003) 110.
Journal of Operations & [11] Torres, A.J.R. and Centeno, G.
Production Management, 15(3) "Minimizing the number of late jobs
(1995) 50-58. for the permutation flowshop
[5] Chan, F.T.S., Wong, T, C. and problem with secondary resources",
Chan, L.Y., "A genetic algorithm- Computers and Operations Research,
based approach to machine 35(4) (2008) 1227- 1249.
assignment problem", International [12] Voss, C., Tsikriktsis, N. and
journal of production research, Frohlich, M., “Case research in
43(12) (2005) 2451-2472. operations management”,
[6] Chan, F. T. S., Wong, T. C. International Journal of Operations &
and Chan, L. Y., "Flexible job-shop Production Management, 22(2)
scheduling problem under resource (2002) 195-219.
constraints", International Journal of [13] Weng, M. and Haiying, R., "Job
Production Research, 44(11) shops to minimize mean tardiness", IIE
(2006) 2071-2089. Transactions, 38(9) (2006) 789-795.
[7] Gupta, J.N.D., Koulamas, C.P.,
Kyparisis, G.J, Potts, C.N. and
Strusevich, V.A., "Scheduling
three-operation jobs in a two-
machine flow shop to minimize
makespan", Annals of Operations
Research, 129(1-4) (2004) 171-
185.
 Table (1) Possible Priority Rules
Rules Description Type of rule
FCFS First Come First Served Local rule
SPT Shortest Processing Time Local rule
EDD Earliest Due Date Local rule
CR Critical Ratio Global rule
S/O Slack per Operation Global rule
RUSH Emergency or Preferred Customer Local or Global
First

Table (2) Setting up Time Matrix

Setup Setup
Time Time
Job Number Conditions Conditions
for for
M1 M2
1 ST11 If Job 1 done ST12 If Job 1 done
first first
2 ST21 If Job 2 ST22 If Job 2 follows
follows Job 1 Job 1
3 ST31 If Job 3 ST32 If Job 3 follows
follows Job 2 Job 2
: : : : :
: : : : :
: : : : :
STN 1 If Job N STN 2 If Job N follows
N follows Job N- Job N-1
1

Table (3) Data for Jobs and Processing Time

Job Processing Time Processing Time
Number (Hr) for Vertical (Hr) for
Cutting Machine Circular Cutting
Machine
1 20 4
2 10 12
3 3 5
4 10 8
5 5 6
6 2 12
7 8 4
8 7 10
9 3 6
10 4 1
Table (4) Job's Setup time (minutes) on Vertical Cutting Machine
Jobs 1 2 3 4 5 6 7 8 9 10
1 5 18 13 24 18 20 15 19 15 1
2 15 18 10 4 17 20 17 25 18 29
3 16 20 13 15 18 19 15 30 10 15
4 10 11 20 16 12 25 3 22 12 8
5 20 28 17 35 28 5 16 6 30 23
6 15 18 10 17 15 2 20 3 5 9
7 2 7 15 28 17 9 11 3 25 8
8 30 6 14 10 25 17 12 13 15 15
9 9 2 17 1 9 9 20 25 30 7
10 9 2 18 15 7 33 14 45 19 33

Table (5)Job's Setup time (minutes) on Circular Cutting Machine

Jobs 1 2 3 4 5 6 7 8 9 10
1 5 25 18 30 15 12 13 10 12 15
2 9 15 20 30 8 33 17 25 14 19
3 17 25 29 25 11 18 25 17 18 27
4 12 30 25 19 28 22 2 7 17 30
5 5 8 16 20 25 15 31 17 17 22
6 7 25 5 14 20 19 10 9 13 2
7 6 1 4 5 10 8 9 12 16 4
8 9 20 21 18 9 12 25 18 7 19
9 10 15 17 13 12 18 20 25 11 32
10 7 3 8 5 17 20 8 45 5 15

Table ( 6) Calculation of Setup time by Existing Sequence

Job M1 M2 Setup Time
T(M1) T(M2)
No. Ti To Ti To ST1 ST2 TST
1 20 4 0 20 20 24 5 5 10
2 10 12 20 30 30 42 18 25 43
3 3 5 30 33 42 47 10 20 30
4 10 8 33 43 47 55 15 25 40
5 5 6 43 48 55 61 12 28 40
6 2 12 48 50 61 73 5 15 20
7 8 4 50 58 73 77 20 10 30
8 7 10 58 65 77 87 3 12 15
9 3 6 65 68 87 93 23 7 30
10 4 1 68 72 93 94 7 20 27
72 68 653 118 167 285
Mean Flow Time = 65.3
 Table (7) Generation of four sequences

S1:
6 3 9 5 8 2 4 1 7 10
S2:
6 9 3 5 8 2 4 1 7 10
S3:
6 3 9 5 8 2 4 7 1 10
S4:
6 9 3 5 8 2 4 7 1 10

Table (8) Calculation of Setup Time for sequence S1

Machine1 Machine2 Setup Time
Job T(M1) T(M2)
Ti To Ti To ST1 ST2 TST
6 2 12 0 2 2 14 15 7 22
3 3 5 2 5 14 19 10 5 15
9 3 6 5 8 19 25 10 18 28
5 5 6 8 13 25 31 9 12 21
8 7 10 13 20 31 41 6 17 23
2 10 12 20 30 41 53 6 20 26
4 10 8 30 40 53 61 4 30 34
1 20 4 40 60 61 65 10 12 22
7 8 4 60 68 68 72 15 13 28
10 4 1 68 72 72 73 8 4 12
454 93 138 231
Mean Flow Time = 45.4

Table (9) Calculation of Setup Time for sequence S2

Machine1 Machine2 Setup Time
Job T(M1) T(M2)
Ti To Ti To ST1 ST2 TST
6 2 12 0 2 2 14 15 7 22
9 3 6 2 5 14 20 5 13 18
3 3 5 5 8 20 25 17 17 34
5 5 6 8 13 25 31 18 11 29
8 7 10 13 20 31 41 6 17 23
2 10 12 20 30 41 53 6 20 26
4 10 8 30 40 53 61 4 30 34
1 20 4 40 60 61 65 10 12 22
7 8 4 60 68 68 72 15 13 28
10 4 1 68 72 72 73 8 4 12
455 104 144 248
Mean Flow Time = 45.5
 Table (10) Calculation of Setup Time for sequence S3
Machine1 Machine2 Setup Time
Job T(M1) T(M2)
Ti To Ti To ST1 ST2 TST
6 2 12 0 2 2 14 15 7 22
3 3 5 2 5 14 19 10 5 15
9 3 6 5 8 19 25 10 18 28
5 5 6 8 13 25 31 9 12 21
8 7 10 13 20 31 41 6 17 23
2 10 12 20 30 41 53 6 20 26
4 10 8 30 40 53 61 4 30 34
7 8 4 40 48 61 65 3 2 5
1 20 4 48 68 68 72 2 6 8
10 4 1 68 72 72 73 1 15 16
454 66 132 198
Mean Flow Time = 45.4

Table (11) Calculation of Setup Time for sequence S4

Machine1 Machine2 Setup Time

Job T(M1) T(M2)
Ti To Ti To ST1 ST2 TST
6 2 12 0 2 2 14 25 7 32
9 3 6 2 5 14 20 5 13 18
3 3 5 5 8 20 25 17 17 34
5 5 6 8 13 25 31 18 11 29
8 7 10 13 20 31 41 6 17 23
2 10 12 20 30 41 53 6 20 26
4 10 8 30 40 53 61 4 30 34
7 8 4 40 48 61 65 3 2 5
1 20 4 48 68 68 72 2 6 8
10 4 1 68 72 72 73 1 15 16
455 87 138 225
Mean Flow Time = 45.5

Table (12) Summary of Calculation of Setup Time

Time Time Setup Setup Total Mean

to to time time Setup Flow
Sequence finish finish for all for all Time Time
all jobs all jobs jobs jobs Z=ST1+S
by M1 by M2 for M1 for M2 T2
S1 72 73 93 138 231 45.4
S2 72 73 104 144 248 45.5
S3 72 73 66 132 198 45.4
S4 72 73 87 138 225 45.5
Existing 72 94 118 167 285 65.3
Sequence
Reduction(%) 0 21 52 35 87 30.4
Figure (1) Gantt Chart for Sequencing S3