ManufacturigCase Study
ManufacturigCase Study
ManufacturigCase Study
CONTENTS
Introduction
..
..
..
..
..
..
..
..
..
..
..
..
..
..
Overview
Manufacturing the Parts
The Trim Operation
Stacking the Products
The Palletizer
Pallets
Conveyors
Reliability
Work Patterns
4
5
6
6
7
7
7
8
8
9
9
10
11
..
..
Modeling Tips
Experimentation Tips
..
..
.
.
12
12
13
13
13
14
18
INTRODUCTION
This case study exercise is designed as an example simulation project conducted in
WITNESS. Students, and others undertaking this exercise, are expected to build a model,
conduct experimentation and try their own ideas for improvement.
The Pallets
The pallets used are expensive custom items (due to the need to keep the parts
particularly rigid in transport). The company currently owns just 12 pallets.
Conveyors
All conveyors in this facility are of the accumulating type in WITNESS terms they are
queuing conveyors all parts will roll up to the queue of parts at the front. The time for all
conveyors to cycle one part position is 0.5 Minutes.
Back to Contents Page
The Trim machine breaks down every 100 to 200 minutes (UNIFORM distribution) and
takes 20 to 25 minutes to repair (UNIFORM distribution)
The Palletizer breaks down every 10 to 20 operations (UNIFORM distribution) and takes
between 1 and 4 hours to repair.
The Trim and the Palletizer machines are repaired by the same person who prioritizes
work on a first come first served basis.
Work Patterns
The factory is run 24 hours a day, 7 days a week. Each position involving people is
manned for the entire duration of production.
EXERCISE 1
Building your Model
Using the information provided, build a WITNESS model of the current facility.
You can use the sample models for guidance, but you are free to design your model in any
style you see fit.
Note: You will notice that this model has some variation in it. You will need to determine
how long a model run is needed before the production settles down to give an accurate
average figure.
10
EXERCISE 2
The factory manager wants you to evaluate implementing each of these options and report
back on the increased average number of parts per week which can be produced.
Produce a clearly labeled table of results for the number of scenarios that you think fit.
Include details of how long the model was run to achieve the demonstrated results and
explain in a short paragraph how the run time was determined. In the results table clearly
include both experimentation parameters and the key results achieved. You may even
include graphical representation of results.
Back to Contents Page
11
EXERCISE 3
Further Analysis & Experimentation
The manager of the factory is delighted with your analysis in Exercise 2 and decides to
give you much greater responsibility for improving the factory.
Finance have determined that each additional part that can be produced per week will
increase revenue by $5.
Overall there is a limit on spend of $50,000 in the next year. All capital costs need to be
depreciated over just one year as this industry changes rapidly. Therefore all costs are
incurred over the year being planned.
Using the available improvement options listed on the next page; identify and produce a
report that indicates how best to invest the improvement budget. Include a clear table of
the result you have found together with your recommendations.
12
Further Options
Hiring extra people (stackers or maintenance people) - the cost over the year for each
new person/position (i.e. at a position 24 hours a day) would be $30,000
Improving the reliability of the trim machine through a better maintenance program
(intervals between breakdowns up by 10%) costs $2,500 per year
Improving the repair times of the palletizer machine by 20% through the use of hired
advanced repair tools costs $3,500
From these options alone identify and report on the best use of the investment money.
13
EXERCISE 4
Zero Spend Options
Often in models such as this one there are options that will cost nothing and yet will
improve the situation. These often include ways that production can be controlled
differently. For example a different priority of selection of work on to a machine, or tasks
for a person carried out in a different priority order. People can even be called away from
one task to complete another if it arises read about Labor pre-emption in the help text to
explore this option.
The manager would like to know are there any options here that would improve the base
scenario with no spend at all?
Identify and evaluate any improvement options that will not result in additional cost for
Steel Nuggets Inc. Report back on any improvements you have found.
Further Ideas
Identify whether you think there might be any additional ways that production could be
improved through the spending of more money.
List any ideas that you have that could have an effect, regardless of the cost implication.
14
MODELING TIPS
The following topics are designed to help you locate further information if need assistance
with modeling particular areas of layout and logic.
Think of elements as your building blocks that have different properties and have different
purposes they do not necessary have to represent what they are called.
Modeling Labor
Even though there is a labor element in WITNESS, it does not mean that this needs to be
used every time a person is involved in a task. Sometimes a simpler alternative is to model
the person as a machine element, staying in a particular location and processing parts just
like a machine.
In this model there is a need for a labor element in one area of the model and there is also
a place where it is simpler to model the person as a machine.
Back to Contents Page
15
MODELING TIPS
Attributes and Variables
There are a number of places in this model where attributes can be used effectively. The
values for the attributes can be set in Actions on Create on the dialog and then the
attribute names can be used in other element dialog fields for example, within the cycle
time of a machine expression.
Variables provide a good place for storing information. There are many uses including the
ability for a value to dictate which logic is chosen in a section of a model for example
which buffer a machine is currently inputting from or outputting to.
Conveyors
All the conveyors in this model are Queuing type and the simplest way to model this type
of conveyor is to select Indexed Queuing from the type drop down box on the Conveyor
detail dialog and then enter how many parts can queue on the conveyor in the Length in
Parts field.
16
MODELING TIPS
Delay Buffers
There are a number of places in this model where delay buffers can be used effectively. A
buffer can have either a maximum or a minimum delay defined (or both). After the
minimum delay a machine (or other element input) rule is allowed to pull the part out of
the buffer. After a maximum delay the output rule on the buffer is run (just once it will
not block and retry).
This is a simple way to model the time for an activity which can accept other parts in to
the activity at the same time as others are being processed.
An example of this might be a delay for parts to cool down after being cooked, or a
transportation delay where a delivery mechanism is not being limited by other resources
(such as vehicles).
Breakdowns
All machine element detail dialogs contain a tab where breakdowns can be defined.
17
MODELING TIPS
Limited Part Arrivals
Where there is a fixed number of a particular part in a model, for example pallets, the part
detail dialog offers a field called maximum arrivals which can be set to stop too many
entering the model.
Rules
A number of different rules are useful in this model. One model that contains a SEQUENCE
rule is Bells.mod accessible from the start page sample models option. In this model
the Assembly machine uses a SEQUENCE input rule.
18
EXPERIMENTATION TIPS
Make sure that you look at the report tables and charts produced by WITNESS carefully in
your evaluation of each scenario. These are accessed by right clicking on a WITNESS
element and selecting the Statistics option.
In this model the machine reports are especially useful. Right clicking on the Simulation
folder of the WITNESS element tree (View -> Element Tree) and clicking on Statistics will
display the reports on all elements within the model (using the >> and << buttons on
the report screens will move between different element type reports).
The status charts provide a graphical view of the statistics which some people find much
easier to use in order to see the key information. For example in the following chart it is
clear that the machine AA is much more likely to be a bottleneck in the operation than
machine BB. AA has almost no idle time being either busy, blocked, broken down and
being repaired or broken down and waiting for labor for repair. All these states give clues
as to what improvements might be possible:
How can the machine AA be unblocked?
Can a labor resource be made more
available for repairing AA?
19
EXPERIMENTATION TIPS
Experimentation in the WITNESS Workbooks
Many of the examples in the two Learning WITNESS books accessed from the WITNESS
Start Page include conducting experimentation and interpreting the reports. Time spent
studying these examples will increase your model analysis skills.
Speed to Answers
To get some idea of which improvements are worth making it is often a good idea to try a
number of shorter model runs which can be accomplished more quickly. This will help
determine which are the obvious good ideas and the poor ideas. The longer runs to
determine the more accurate answers then need only be carried out for the better
solutions.