CHI 97 Electronic Publications: Late-Breaking/Short Talks
Interfaces for Advanced Manufacturing Technology
Michelle Vazquez and Marc L. Resnick, Ph.D.
Industrial and Systems Engineering
Florida International University
Miami, FL 33199, USA
+1-305-348-3537
resnick@eng.fiu.edu
ABSTRACT
Due to the rapid computerization of advanced manufacturing
workplaces, there is an increasing need for interfaces which can
support this specific set of applications and users. However,
workers in these situations tend to be highly trained in the specific
tasks which they must accomplish, but may be relative novices
when it comes to using computing systems. This paper describes
the design of Easy Assemble, a windows based support
system to assist workers in a flexible assembly task. Six subjects
used Easy Assemble as real-time instructional support to
assemble four products in a simulated manufacturing environment.
Subjects assembled products in less than half the time and with
variances much lower than the control group which used the traditional
method of blueprints. Furthermore, subjects made significantly
fewer errors. The system provides a starting point for the development
of fully integrated systems for the advanced manufacturing environment.
Keywords
Advanced manufacturing, novice, computer-based training
© 1997 Copyright on this material is held by the authors.
INTRODUCTION
In order for manufacturers to maintain a competitive
advantage in today's market they must provide customers with flexibility
(Noaker 1994). They need to provide products having a customized
combination of features in short lead times. This has eliminated
the traditional assembly line where workers produce one product
continuously. Facilities must have the ability to shift assembly
processes instantaneously. In order for low- skilled workers to
be able to make this transition to new products without significant
off-line training, an effective on-line support system is needed.
This system should reduce training requirements without increasing
error rates or introducing quality problems.
This study tested an on-line system designed using
graphical user interface (GUI) standards developed for computer-assisted
instruction, hypermedia and on-line help (Grabinger, 1993). The
assembly instructions followed the guidelines of Dixon (1988)
for step by step instructions.
METHODS
Subjects
Six subjects were recruited for the experiment from
the general population in Miami, FL. Subjects needed to be fluent
in English and have no previous experience with computers in order
to simulate the experience of current assembly workers. The age
of the subjects ranged from 25 to 57 years old with an average
of 40.
Apparatus
The four products used to test the proposed system
were simple constructions created using MegaBlocksTM.
Two different sets of instruction support materials were used
during the experiment. The first set consisted of standard blueprints
(Engstrom et al 1994) for each of four different products. The
blueprints provided all the information related to the physical
configuration of the final part. This information included dimensions,
components, and different views. The second set of materials was
Easy Assemble, the proposed real-time computer support
system. Easy Assemble provided interactive assembly instructions
through on-line directions and illustrations in a GUI-based system.
Easy Assemble interface
This system was developed using the standard edition
of Microsoft Visual Basic®, Version 4.0. Some of the features
of the application are:
- The operator can select the product to be assembled
by clicking with a mouse-button on a picture of the product.
- The operator, at any desired time, can select
a full-screen picture of all the assembly components required
by the final product.
- The operator, at any desired time, can select
a full-screen picture of the completed assemblage.
- The system provides directions and illustrations
on assembly operations, in their executable order, at a pace controlled
by the operator.
- The system repeats instructions if requested.
- A two-button mouse is the only input device needed.
Procedure
Training
In order to become familiar with Easy Assemble,
subjects were provided with a training session during which
they were shown how to operate the mouse and the control buttons
to navigate through the application. The training required an
average of only ten minutes for each subject.
Data Collection
The order in which the subjects assembled the products
was rotated. Subjects used both instruction materials, half using
the blueprints first and half using Easy Assemble first.
In each segment, the subject was provided with all the assembly
components for the four products, but assembled only two of them.
The subjects were instructed to be as efficient and accurate as
possible, assuming the following: (1) he/she was paid by the number
of products completed, and (2) if the final product contained
an error in the assemblage, it will be discounted from the total
number completed at the time of payment. An error in the final
product was considered to be any of the following: (1) a component
in the wrong orientation, (2) a component in the wrong location,
(3) a missing component, or (4) a component not belonging to the
final product. However, if the subject made one of those mistakes,
he/she was not notified until the end of the experiment. After
the first segment was completed, the products were disassembled
and again placed before the subject. The instructional materials
were switched and the subject was required to assemble the other
two products under the same conditions. The time to complete each
individual assembly and the total number of errors present in
each final product were recorded during both segments.
RESULTS
The average assembly times and errors are shown in
Table 1. The two systems were compared using an Analysis of Variance
(ANOVA). Even though subjects lacked experience using computers,
the assembly time was still improved when using Easy Assemble
compared to the times when using the blueprints. In all cases,
the average assembly times when using Easy Assemble were
significantly shorter than when using the blueprints (p < 0.05).
Another significant result is that assembly times varied significantly
among the product types when the blueprints were used, but not
with Easy Assemble. Finally, subjects made significantly
fewer errors when using Easy Assemble compared to the blueprints
(p < 0.05).
DISCUSSION
The results of this study have some significant implications
for the design of interactive support systems for advanced manufacturing.
The inclusion of a properly designed interactive system can improve
the performance of assembly workers who may be experts in the
assembly process but have very little experience in the use of
computers. The improvements noted in this study were threefold.
First, subjects were able to assemble the products faster with
Easy Assemble than with traditional blueprints. Obviously,
improved speed of assembly will increase the productivity of the
workplace. Secondly, the reduced variance of the assembly task
can improve the ability of industrial engineers in the scheduling
of other tasks in the manufacturing process. This enhances the
ability of manufacturers to take advantage of Just-In-Time Inventory
control systems, which can decrease many other costs in the workplace,
such as inventory and material handling. Finally, subjects made
fewer errors when using Easy Assemble. Fewer errors will
increase the quality of the resulting products, thus decreasing
many costs in the workplace, such as those associated with rework,
scrap, recalls and warranties. Higher levels of quality can also
increase a company's reputation, leading to higher sales in the
future. Thus the potential of implementing interactive assembly
instructional systems in the workplace can lead to significant
improvements in the profitability of companies using advanced
manufacturing.
REFERENCES
Dixon, P. (1987). The processing of organizational
and component step information in written directions. Journal
of Memory and Language, 26, 24-35.
Engstrom, T., Medbo, L., and Jonsson, D. (1994).
Extended work cycle assembly - A crucial learning experience.
International Journal of Human Factors in Manufacturing,
4(3), 293-303.
Grabinger, R. S. (1993). Computer screen designs:
viewer judgments. Educational Technology Research and Development,
41(2), 35-73.
Noaker, P.M. (1994). The search for agile manufacturing.
Manufacturing Engineering, 40-43.
Table 1. Assembly Times and Errors for subjects
using Easy Assemble and the standard blueprints
| Easy Assemble
| blueprints |
| Product | assembly time (mean)
| assembly time (variance) |
assembly errors | assembly time (mean)
| assembly time (variance) |
assembly errors |
| A | 4.34 min
| 1.32 min | 0
| 8.75 min | 1.02 min
| 2 |
| B | 3.81 min
| 0.31 min | 0
| 10.75 min | 1.13 min
| 8 |
| C | 3.74 min
| 0.17 min | 0
| 13.02 min | 0.81 min
| 6 |
| D | 3.25 min
| 0.21 min | 1
| 7.34 min | 1.69 min
| 3 |
CHI 97 Electronic Publications: Late-Breaking/Short Talks
