AC 2010-510: CASE STUDIES FOR LEARNING AUTOMATED SYSTEM

INTEGRATION

Sheng-Jen Hsieh, Texas A&M University

Dr. Sheng-Jen (“Tony”) Hsieh is an Associate Professor in the Dwight Look College of
Engineering at Texas A&M University. He holds a joint appointment with the Department of
Engineering Technology and the Department of Mechanical Engineering. His research interests
include engineering education, cognitive task analysis, automation, robotics and control,
intelligent manufacturing system design, and micro/nano manufacturing. He is also the Director
of the Rockwell Automation laboratory at Texas A&M University, a state-of-the-art facility for
education and research in the areas of automation, control, and automated system integration.

Page 15.264.1

© American Society for Engineering Education, 2010

 Case Studies for Learning Automated System Integration
Abstract

Research indicates that the use of case studies for learning engineering design has a positive
impact on generating excitement about engineering, conveying industry “best” practices, and
demonstrating the design process. In addition, exposure to relevant cases can help in problem-
solving and automated system design. Interviews with automated system integrators have shown
that they often recall other systems they have seen or worked on previously—that is, cases—in
coming up with conceptual designs.

This paper will describe the development of case studies to help students to learn the automated
system design process. The case studies are based on examples from industry and illustrate good
industrial design practice. Each case study walks learners through the stages of coming up with
a conceptual design, including: 1) identify requirements; 2) collect data; 3) determine product
assembly sequence and cycle time; 4) determine equipment required for assembly process; 5)
determine layout of assembly line; and 6) perform cost estimation and analysis

Motive

Automation has a profound effect on the way we do work. A U.S. Census Bureau report notes
that yearly exports in the flexible manufacturing category (equivalent to industrial automation)
were $19.44B in 2006, a 10% jump from $17.61B in 20051. Moreover, monthly exports in the
flexible manufacturing category were $4.06B in March 2008, a 0.5% jump from $4.04B in
March 20072. This trend is likely to continue to increase as the manufacturing sector continues
to transform to a high tech, less labor-intensive and value added industry using advanced
automated systems.

Integrating the components of an automated manufacturing system requires knowledge about the
various mechanical and electrical devices available to make up the system—including their
functions, power requirements, and specific characteristics—and the ability to write PLC
programs to orchestrate and synchronize the process being automated. It is a complex cognitive
skill, and often there is no course available that teaches it. In addition, some colleges do not
have the equipment resources needed to provide hands-on experience with automated systems.
Consequently, new automation and control engineers are often not fully prepared to perform
system integration tasks. Needed are readily available instructional materials that can better
prepare new engineers for these challenging tasks.

A study by Hsi and Agogino3 suggests that the use of case studies for learning engineering
design has a positive impact on generating excitement about engineering, conveying industry
“best” practices, and demonstrating the design process. Hsieh4 has noted that automated system
integrators often recall other systems they have seen or worked on previously (i.e., cases) in
coming up with conceptual designs. Although there are quite a few programs that depict
manufacturing processes (notably the cable TV show “Made in America”), there are relatively
few instructional materials that systematically walk learners through the process of designing an
Page 15.264.2
automated system. In this paper, we present the process of the developing such a case study and
illustrate the principles with a real-life example.

Case Study Development Process

Over 40 engineers at system integration companies in the U.S. and Europe were interviewed and
asked to describe how they would go about designing a new automated manufacturing system.
The interviews were transcribed and analyzed. The results and theoretical work were evaluated
to come up with a generalized process for automated system design. This process was used to
guide the case study development. Steps include:

1. Identify requirements.
2. Collect data.
3. Determine assembly sequence and cycle time.
4. Select assembly line components.
5. Determine layout of assembly line.
6. Perform cost estimation and analysis.

This process described in detail in the following sections and illustrated with screen shots from a
case study on how to design an automated system for crayon manufacturing. The actual case
study can be found at: http://etidweb.tamu.edu/hsieh/Autosys/flash/Case_studies_learning.html

1. Identify requirements

Often the customer will provide a product part list, available budget, desired production rate,
delivery time, system up-time, and space constraints. Figure 1 is an example of customer
requirements from the crayon manufacturing case study. For a system integrator, the main focus
is to understand the requirements from the customers in as much detail as possible. Common
requirements include cost, production rate, system up time, and delivery time. The proposed
conceptual design in the proposal should transform these needs into concrete design features that
achieve the requirements.

Quality Function Deployment (QFD) is a method originally developed by the designers of

Japanese automobiles to prioritize needs into concrete actions and metrics. Figure 2 shows the
QFD “House of Quality.” The customer requirements are written on the left square with the label
of ‘WHAT.” A rating of the customer requirements is also listed in the same square. Based on
these, system integrators can create columns that list “how” the customer requirements might be
satisfied and measurements needed to access the technical requirements. The square on the right
is used to document customer perceptions of competing products. Figure 3 is an illustration of
QFD applied for a product (a kitchen broom). The needs are translated into technical
requirements with process parameter target values to be controlled.
Page 15.264.3
 Figure 1. System requirements

Figure 2. Quality Function Deployment Figure 3. Quality Function Deployment Example

2. Collect data

An application engineer at a system integration company will want to collect as much detail as
possible about the part list, process requirements, and how parts will be presented to the
assembly line. For example, will the parts be delivered in a tray or a box? Will they come from
another sub assembly line or be shipped from a vendor? Is a machine needed to check the parts
before assembly? How big is a lot? How often do you need to refill the part feeder? Are there
any pre-existing machines that need to be incorporated into the line?

3. Determine the Assembly Sequence and Cycle time

Given the required production rate, system up-time and part list, an optimum
assembly/manufacturing sequence can be designed. Figure 4 shows the crayon manufacturing
process. A precedence diagram (shown in Figure 5) will be constructed based on operations
precedence relationships; then a line balancing method, such as Largest-candidate rule, will be
employed to combine work elements into a workstation, and the original cycle time will be
Page 15.264.4

calculated. After that, balance delay can be calculated to measure line inefficiency and the
original cycle time will be compared with the desired cycle time derived from the predicted
production rate. Parallel workstations and/or multiple identical production lines will be arranged
to reduce default cycle time and/or reach the desired production rate. Figure 5 shows parallel
workstation and parallel lines arrangements and Figure 6 shows the calculation of cycle time.
The desired workstation arrangement is for each station to have the same operation time and the
line inefficiency to be zero.

Figure 4. Major steps in crayon Figure 5. Precedence diagram for

manufacturing process design alternative A

Figure 6. Cycle time calculation

4. Select assembly line components

This task involves making decisions about (1) the number of machines for each station; (2) the
types of machines and machine capacity for each station; (3) types of material handling and
transfer equipment needed, (4) types of part feeders and machine tools required, (5) number of
operators, and (6) number of work shifts needed to manage the assembly lines. All these factors
will affect the overall cost of the proposed design. Therefore, a few alternatives may be
Page 15.264.5
proposed. For example, in the crayon manufacturing case study, to achieve the desired
production rate of 14.2 million units per day, the design incorporates parallel stations/machines.

Figure 7. Calculating amount of

equipment needed.

5. Determine layout of assembly line

There are three common line layouts: inline, U-shape and parallel cell configurations. The goal
is to reduce material transfer time (since transfer is not a value-added process) and to increase the
feedback. The crayon manufacturing system incorporates 13 to 16 parallel lines (illustrated in
Figure 8). The inline layout allows raw materials to be placed at one end and finished product at
the other.

Page 15.264.6

Figure 8. Proposed inline system layout.

6. Perform cost estimation and analysis

Cost categories include labor cost, machine/equipment cost, materials cost, system development
cost, and overhead. Several alternatives may be proposed. Customers may choose an alternative
based on its rate of return, request a hybrid of two or more of the proposed alternatives, or not
accept any proposed design. In the crayon manufacturing case study, two alternatives are
proposed (Figure 9). One alternative is to have fewer parallel workstations and parallel
production lines but work three shifts (Design A) versus more parallel workstations and
production lines but work two shifts (Design B).

Figure 10 shows the molding process from the proposed conceptual design and Figure 11 shows
that conceptual design A was favorable due to its higher rate of return given the same unit profit.

Figure 9. Cost comparison of designs A and B

Figure 10. Conceptual design - crayon molding Figure11. Customer selection.

Page 15.264.7
Conclusion and Future Directions

The crayon manufacturing case study is an example of a system for a discrete manufacturing
process. Future case studies will also focus on systems for continuous processes (such as oil
refining) and for hybrid processes (such as beverage manufacturing, and also on machine and
system design. Case studies may also incorporate other steps in the design process, such as
design analysis (including cost analysis) and ranking; prototyping; and field test/engineering
changes.

Acknowledgements

This material was supported by a National Science Foundation grant no. 0837634. Any
opinions, findings, and conclusions or recommendations expressed in this material are those of
the author and do not necessarily reflect the views of the National Science Foundation.

Bibliography

1. Online resource available at:

http://www.census.gov/foreign-trade/statistics/product/atp/2006/12/atpctry/atpg06.html (last accessed on March
27, 2010)
2. Online resource available at:
http://www.census.gov/foreign-trade/statistics/product/atp/2008/03/atpctry/atpg06.html (last accessed on March
27, 2010)
3. Hsi, S. and Agogino, A.M. “The impact and instructional benefit of using multimedia case studies to teach
engineering design,” Journal of Educational Multimedia and Hypermedia, v 3, n 3-4, 1994, p 351-76
4. Hsieh, S. "Automated Manufacturing System Integration Education: Current Status and Future Directions,"
Proceedings of the 2005 ASEE Annual Conference & Exposition, Portland, OR.

Page 15.264.8