1.

Problem Description
Consider a manufacturing industry producing components for various automotive
manufactures.
Manufacturing industry, from the unit view, can broadly be divided into three main
categories namely
1. Transmission Line,
2. Engine line and
3. Component Section.
Thus for operations overview for only Transmission line was considered. Further, the starting
point from which material enters the line is Soft cell. Majority of the parts travels through
Heat treatment, Sleeve cell and hard cell only after passing from Soft cell. Thus on
transmission line, focus will be more on soft cell

Figure 1: Overview

A cellular layout is used at the shop floor. Each of the cells produces a job family consisting
of multiple job types.
Each job type has certain demand. Demand for each month varies. Let us consider, a
representative cell called ‘SL’ which processes 11 different jobs: J1, J2…, J11.
These jobs undergo 49 machining operations. These operations are carried out on 23 non-
identical machines (M1, M2, … , M23).
The process flow for each job is predefined by the process engineers and is shown in Table 1.
For example, the first operation of job 1 (i.e., O1) is scheduled on machine M1, the second
operation (i.e., O2) is scheduled on machine M3 and the third operation (i.e., O3) is scheduled
on machine M6. Thus, job 1 (i.e., J1) is processed from machine M1, M3 and M6 only.
Similarly, the flow of all the jobs in different machines can be read from Table 1.
 Table 1 Process flow of jobs
Ox/Ji O1 O2 O3 O4 O5 O6 O7 O8 O9
J1 M1 M3 M6 - - - - - -
J2 M1 M3 M6 - - - - - -
J3 M1 M3 M5 - - - - - -
J4 M1 M3 M5 - - - - - -
J5 M14 M3 M5 - - - - - -
J6 M14 M13 M18 - - - - - -
J7 M14 M6 - - - - - - -
J8 M4 M14 M23 M5 - - - - -
J9 M7 M8 M9 M12 M21 M22 M17 M10 M19
J10 M4 M20 M13 M22 M18 M19 - - -
J11 M9 M2 M16 M11 M13 M15 M18 M19 -

One example process flow for job J9 is as described in figure 2. The sequence of these jobs on
various machines and their batch-sizes are currently decided by production manager and
supervisors based on their experience. The initial batch size and initial sequence of jobs in the
respective machine for each operation are presented in Table 2. Also, in Table 2, jobs
properties i.e. processing times (𝑃𝑇) setup times (𝑆𝑇), are offered.

Figure 2: Representative process flow of job J9

To summarize following are the problem specifications:

• No. of machine: 23
• No of job types: 11
• Total number of operations: 49
• Initial demand: 3000 for each job type
• Planning horizon one month (i.e. 43200 minutes)
• Machine characteristics:
▪ Processing time for each job
▪ Setup time for changeover of job
▪ Failure distribution and repair distribution
▪ Each machine is equipped with buffers. The buffers hold raw material or work
in process or semi-finished or finished items.
 The production of jobs depends on the availability of machines. The time to failures
of the machines follow a two-parameter Weibull distribution. Shape parameter 𝛽𝑗 and
scale parameter Ƞ𝑗 of these distributions are shown in Table 2. Repair time is not
constant and is distributed according to lognormal distribution for each machine with
lognormal (1800 minutes, 600 minutes).

Machine (𝜷𝒋 and Batch Size

Operation- Sequence
𝑷𝑻(Minutes) 𝑺𝑻(Minutes) of each
Ƞ𝒋 (𝒊𝒏 𝒎𝒊𝒃𝒖𝒕𝒆𝒔)) Job Number
job
M1 (2.3, 42000) O1-J1 1.57 90 300 1
O1-J2 1.4 90 300 2
O1-J3 1.3 90 300 4
O1-J4 1.8 90 300 3
M2 (2.7, 30000) O2-J11 4.5 0 300 -
M3 (2, 36000) O2-J5 0.6 20 300 1
O2-J2 0.6 20 300 2
O2-J1 0.6 20 300 4
O2-J3 0.6 20 300 3
O2-J4 0.6 20 300 5
M4 (2, 48000) O1-J10 1.6 90 300 2
O1-J8 1.8 420 300 1
M5 (3, 32400) O3-J3 1.1 60 300 1
O3-J4 1.24 60 300 3
O3-J5 1.1 60 300 4
O4-J8 1.1 60 300 2
M6 (2.6, 25200) O3-J1 1.8 60 300 2
O2-J7 1.9 60 300 3
O3-J2 2.1 60 300 1
M7 (2.6, 28800) O1-J9 7.6 60 300 -
M8 (1.9, 33000) O2-J9 8 60 300 -
M9 (2, 48000) O3-J9 2.9 60 300 2
O1-J11 1.2 30 300 1
M10 (2.4, 42000) O8-J9 2.66 0 300 -
M11 (2.3, 30000) O4-J11 8.1 0 300 -
M12 (2.9, 37200) O4-J9 2.23 0 300 -
M13 (1.8, 27600) O5-J11 1.2 60 300 1
O3-J10 0.6 30 300 3
O2-J6 0.6 30 300 2
M14 (1.3, 27000) O1-J5 1.3 90 300 1
O1-J6 1.4 90 300 4
O2-J8 1.5 90 300 2
O1-J7 2.3 90 300 3
M15 (2, 48000) O6-J11 4.3 180 300 -
M16 (2.6, 42000) O3-J11 10.6 60 300 -
M17 (2.1, 51000) O7-J9 1.86 45 300 -
M18 (2.5, 42000) O7-J11 1.4 60 300 3
O5-J10 1.3 60 300 1
O3-J6 1.4 60 300 2
M19 (2.4, 30000) O9-J9 0.56 90 300 1
O8-J11 0.56 90 300 2
O6-J10 0.6 90 300 3
M20 (2.3, 36000) O2-J10 7 180 300 -
M21 (1.5, 22800) O5-J9 1.2 30 300 -
M22 (1.2, 24000) O6-J9 1.3 45 300 2
O4-J10 1.6 45 300 1
M23 (2.7, 36000) O3-J8 2.12 0 300 -

An optimized job sequence of production is important for timely completion and fulfilling
the demand of components.
The jobs are processed in batches having several items in a batch. Batch-sizing is another
key decision, as batches having large size will reduce the number of setups but will increase
the WIP inventory level.
Identification and optimization of right decision variables is critical. Few of the decision
variables are described below:
Sequence:
There are many possible sequences for processing. For example, there are four jobs to be
processed on M1. They can be sequenced in 24 different ways. Similarly, M3 processes five
jobs so there are 120 possible ways to prepare sequences for M3. Both M1 and M3 together
would form 2880 of permutations. When combined for all the machines together billions
(1.72 × 1010 ) of possible permutations emerge.
Batch size:
Jobs are processed in batches. One-time setup would be required to process a batch. But
larger batch may create undesired idle time. So, an optimum batch size would be required
such that throughput is maximum.
Preventive maintenance:
Machines are subject to failures. After certain cycles of the machine use, it may fail more
frequently. Some machine may fail more frequently than other machines. Optimizing
preventative maintenance schedule these machines is another important decision on any shop
floor.
Other decisions like inventory level, resource optimization, capacity decision etc. are also
involved in any such shop floor.
The above industrial case is a good example to showcase the value of simulation modelling in
understanding the shop floor dynamics and identifying bottlenecks.

2. Modelling and Simulation

Model the industrial scenario mentioned above in problem description on Tecnomatix
Plant Simulation platform.
The model is built by bringing in machines, jobs and buffer on platform. Jobs are detailed
with job characteristics. Machines are detailed with machine characteristics. The right flow of
each job is given such that all jobs take right path as described by manufacturing scenario.
Following figure gives complete overview of the model.

Figure 3: Overview of the model (Representation)