reading and writing

HUMAN PERFORMANCE MEASUREMENT, MODELING AND SIMULATION FOR AN ASSEMBLY TASK

BY

POONAM LAXMAN DESHMUKH, B.E.

ABSTRACT

HUMAN PERFORMANCE MEASUREMENT, MODELING AND SIMULATION FOR AN ASSEMBLY TASK

BY

POONAM DESHMUKH, B.E.

Master of Science in Industrial Engineering (major) and

Electrical & Computer Engineering (minor)

New Mexico State University, Las Cruces, New Mexico, USA

The primary objective of this project is to measure, model and simulate the human/operator performance in a manufacturing cell to improve the decision making process of the managers. It is well known that people working in a manufacturing facility suffer from stress, fatigue and physical exhaustion due to repetitive manual labor. The purpose of this project is to identify and measure the performance metrics that affect the worker’s performance and help in making decisions about rotating the workers in such a way that their capability matches the task requirement. The project involved, conducting a pilot study to identify the metric of operator performance, physically modeling and simulating an assembly station of a manufacturing cell in a laboratory, measuring the identified metric (dexterity) in the simulated and real environment and compare the results from both the environments to evaluate the simulated assembly station. Using the simulated assembly station, measurements of several different metrics can be performed in future. The primary outcome of this project is the operator task capability-requirement matrix for the assembly station in terms of dexterity. The secondary outcome of this project is the evaluation of the simulated assembly station using t - student test.

Keywords: Human performance measurement, dexterity, manufacturing cell, operator performance measurement, modeling and simulation.

TABLE OF CONTENTS

TOPICS Page

1. INTRODUCTION 12

1.1. Metric Identification 12

1.2. Measurement 14

1.3. Modeling 14

1.4. Simulation 16

2. RELATED RESEARCH 17

3. METHODOLOGY 24

3.1. Pilot Study 24

3.1.1. Equipment and Software 24

3.1.2. Experiment Design 24

3.1.3. Data analysis and plots 25

3.2. Simulation 27

3.2.1. Equipment 27

3.2.2. Experiment Design 28

3.2.3. Data analysis and plots 29

3.3. Main Study 29

3.3.1. Equipment 29

3.3.2. Experiment Design 29

3.3.3. Data analysis and plots 29

4. RESULTS 29

5. DISCUSSION 29

6. CONCLUSION 29

APPENDICES

A. Operator Consent Form 29

B. Manager Consent Form 29

REFERENCES 29

LIST OF FIGURES

Figures Page

1 Fish Bone Diagram 13

2 Anatomy of Hand 15

3 Task Requirement - Capability Model 16

4 (a) Human Glove 23

4 (b) Biomechanics Sensor Glove 23

5 (a) Average reactions Time Plot 25

5 (b) Concentration Plot 25

6 (a) Purdue Pegboard 28

6 (b) Hand - Tool Dexterity Test Equipment 28

INTRODUCTION

It is well known that human performance degrades with repetitive work and time. Repeatability of an assembly task in a manufacturing cell is one of the stressful tasks for a worker. The assembly task can be analyzed to determine the time when an operator gets tired and has to be rotated to some other station for improving the overall cell performance. Operator performance can be analyzed through four main phases - human performance metric identification, measurement, modeling and simulation. There are several causes that can affect the human performance as shown in figure 1.

Figure 1: Fish Bone Diagram (Cause and Effect Diagram) for

Operator Performance in a Manufacturing Cell

Small despcription of the project

Metric Identification:

A metric can be defined as an important aspect to focus on that can be measured over a period of time to communicate vital information for a given situation. There exist several human performance metrics that can be measured either by using sensors or test equipment or software. For identifying the human performance metric we conducted a pilot study and surveyed the workers working in a manufacturing cell at Johnson Controls, Juarez, Mexico. For this project, we identified various metrics that fit for an assembly task:

· Range of motion (if an operator has to stretch for grabbing parts of assemblies),

· Hand-eye coordination (if an operator has to see a drawing and then perform assembly)

· Two-arm coordination (if an operator has to use both hands equally for the assembly), posture (if operator is standing or sitting for a long time)

· Memory (if an operator has to remember a difficult or varying sequence of assembly parts)

· Concentration and attention (if operator surroundings are disturbing)

· Reaction time (if more assemblies have to be performed in less time)

· Manual dexterity (if operator finger movements have to be agile or fast

Manual dexterity is identified as the most important metric among all the other metrics identified. Manual dexterity is defined as the ability to quickly move hand, hand together with arm, or two hands to grasp or to make precisely coordinated movements of the fingers to manipulate, or assemble objects during repetitive tasks. There are two main types of manual dexterity - fine dexterity and gross dexterity. Fine dexterity refers to the ability to manipulate objects using the distal parts of the fingers. Gross manual dexterity or simply manual dexterity involves less refined and less precise movements of the hand and fingers.

Measurement:

Sensors and goniometers that measure real time data can be developed for range of motion and posture. Test equipment can be developed for response speed, dexterity and repeatability measurement. Software modules can be written for concentration, memory and attention measurement. Data gloves that consist of sensors for measuring range of bend of the finger joints can be used to measure dexterity. For measuring dexterity, we used Purdue Pegboard test and Hand - Tool dexterity test equipment.

Modeling:

Operator in a manufacturing cell can be considered as an interactive system. Mital et al. (1993) described several operators modeling approaches to quantify the relationships between the imposed stresses and resulting strain. These approaches are:

Epidemiological: Epidemiology is concerned with discerning the injury patterns present in groups of people doing similar tasks and using these patterns to predict the occurrence of injury. The main task in epidemiological modeling is gathering good physical histories of individuals being considered. The model is designed on the basis of past and present observations and histories to predict future injuries.

Biomechanical: In this model, human body is treated as a system of links and connecting joints and each of the links is the same length and possesses the same mass and moment of inertia as their corresponding human segments. This approach relies on compression and shear forces of the spinal cord and pressures generated in abdominal cavity. Some of the measuring methods include Goniometer for range of motion. Figure 2 shows the anatomy of a human hand.

Physiological: During repetitive handling tasks, a worker’s work capacity is limited by the capacity of oxygen and nutrients being delivered to the tissues and muscles. Some of the physiological measurements include heart rate, blood pressure, and metabolic energy expenditure. The design criterion is the metabolic energy expenditure. Some of the measuring methods include heart rate monitor, BP apparatus, and calorimeter. This approach is based on the capacity-requirement principle.

Figure 2: Anatomy of Hand

Psychophysical: In this model, operators adjust their workload to the maximum amount they can sustain without undue strain or discomfort and without becoming unusually tired. The design criteria are maximum acceptable frequency of handling and maximum acceptable weight/force of handling. Some of the measuring methods are weight lifting, strength testing.

Environmental: In this model, operators get affected by their surroundings. Lighting, temperature, pressure, noise, manufacturing cell layout, placement of workbenches and machines, mood are some of the affecting factors. But as all the workers are working in the same environment, most of these factors are constant with respect to the workers.

Task Requirement - Capability: In this model, the managers try their best to pull up the worker’s task capability towards the task requirement. The task requirement is generally a productivity chart with goals set. Task requirement is the number of assemblies that “should” be performed by a worker in a week, and the task capability is the number of assemblies that a worker “can” perform in a week. Task requirement curve is plotted before the week starts and the job capability curve is plotted after the week ends and both curves can be analyzed for enhancing the worker’s capability for following weeks. As the margin between the task requirement and task capability of a worker decreases, the stress on the worker increases and his or her performance degrades gradually as shown in figure 3.

Figure 3: Task Requirement - Capability Model

Simulation:

An assembly station similar to the one at Johnson Controls was emulated in the Human Performance Laboratory at New Mexico State University. This assembly station was then physically simulated so that there was no need of going to Johnson Controls every time for data collection in future if the simulated and real assembly stations were not significantly different. The emulation consisted of creating an assembly station by considering the lighting system, anti-fatigue mat, workbench, sound, and temperature as in the manufacturing cell A19/A1 at Johnson Controls. The simulation of the assembly station consisted of assembling the parts by the subject in the lab as assembled by Deysey (the worker at the assembly station in Johnson Controls). The dexterity of subject and Deysey was then compared to analyze the difference between the two assembly stations to evaluate the simulated environment against the real environment.

Describe a reason for having literature review road map

2. RELATED RESEARCH

Kondraske (1995a) proposed an elemental resource model (ERM) that interfaces the tasks and the humans through demand and availability of resources. ERM leads to the development of instruments that measure human performance for several different purposes (decision-making, manufacturing, medical, rehabilitation progress, prediction). The ERM model for human-task interface consists of a human model and the tasks imposed on the pool of BEP (basic elements of performance). Entire human in ERM is modeled as a pool of elemental performance resources grouped into four domains (life sustaining, environmental interface, central processing, and information), first three of which are the physical systems (or functional units) while the fourth one is an information base. A human is represented as a set of available performance resources [RAij (t) | Q], where i is the dimension of performance and j is a functional unit, A is available and Q is the operating point. There are three hierarchically levels of a given task - basic element level, generic intermediate level and the high level. A task is modeled in terms of demands imposed on elemental resources and is represented as a set of performance resource demands [RDij (t) | Q], where D is demand. A task is successful when RAij >= RDij inequality is satisfied i.e. available resources are greater than (are enough for) the resource demands for a given task.

Kondraske (1995b, 1995c) and Kondraske and Khoury (1992) derived a major component of the theoretical basis for the GSPT (General Systems Performance Theory) workload model. Using these constructs and equating workload to "stress" as defined in GSPT, any human workload component (W) can be defined as the fraction of available performance resources utilized during task execution:

W (%) = (RDi/RAi) x 100

where RDi = amount of resource of type i demanded (i.e., utilized) and RAi = amount of resource of type i available. This approach allows, by multiplying together the workload quantities computed along separate performance dimensions, a conceptually straight-forward method of representing workload multi-dimensionally. Further, it provides the capability to multiplicatively combine quantities representing workload of subtasks or task components to derive composite measures for predicting workload for upper-level tasks.

Abdel-Malek et al. (2000) discussed the metrics of human performance and modeled them mathematically. The analytically modeled mathematical equations can be implemented in computer programs and by applying the data collected, from several experiments, human performance can be predicted and used for decision-making processes. The measures or metrics discussed include reachability, dexterity, joint functionality, ranges of motion, effort, energy, force, work and power. The reachability can be measured by the volume enclosed by the reach envelope and the range of motion. Dexterity or orientability or manipulability is the degree of possible orientations of an arm at a given target. The effort needed to reach a destination is the displacement required for each joint from initial position to the task completion. Energy is measured in terms of the energy exerted or required by an operator to perform a task. These metrics are useful to model and evaluate human performance so as to automate the ergonomic design process.

Abdel-Malek et al. (2005) presented a mathematical and biomechanical model to solve the placement problem in an ergonomic design. The placement of workers in an assembly line should enable them to maximize dexterity, reach, and minimize stress on joints for designing an assembly line ergonomically. The ergonomic design process is an optimization problem with many variables, dexterity being the major metric or the driving cost function in a manufacturing environment. The objective function is given by f = Dexterity (w), where w (design variables) characterize the position and orientation of the operator and f (design function) is a quantifying measure for dexterity. As the human joints are constrained, every joint can be characterized by an inequality constraint. The hand was modeled by considering the upper arm, lower arm and hand as three consecutive links and then analyzing the rotation of each link with respect to the coordinate system of other two links.

Human performance Technology is an engineering approach to attain the desired accomplishments through human performers. The HPT approach focuses on three issues - problems, opportunities and new situations. A model known as ADDIE (analysis, design, development, implementation and evaluation) was developed for teaching and learning purposes. The HP technologists follow a stepwise method as in ADDIE to implement a proper treatment for any problem. First step consists of the problem definition which is the initiation of a project. Second step is of analysis in which goals, techniques and tools are identified. The analysis of a project is carried on the organization level, the process level and the job or performer level through interviews, observations, surveys, and focus groups. Third step is of design and developments in which a plan of action, design strategies, process developments are the points of focus. Fourth step is of implementation and maintenance in which the project is actually built or implemented and undergoes several modifications according to the changes in the design. Fifth step is of evaluation in which the developed system is verified for its expected operation.

Pennathur et al. (2003) reported the results from an experimental pilot study performed to quantify the manual dexterity of older Mexican American adults. The Purdue pegboard test, a two-arm coordination test, and a hand-tool dexterity test were used in this study. The metrics identified were two-arm coordination and hand-tool dexterity. Purdue pegboard test was intended to measure two types of activities: (1) gross movements of the hands, fingers and arms and (2) finger dexterity, which can be considered as the ability to integrate speed and precision with finely controlled discrete movements of the finger. The two-arm coordination test was intended as a measure of the ability to move both arms in a simultaneous and coordinated manner. The movement involved was that of the whole arm related to the ability of operating-controlling and driving-operating. Hand-tool dexterity test was used to measure the dexterity of participants when using common hand tools. The hand-tool dexterity test was used to measure the manipulative skill, independent of intellectual factors. A t-Student test was conducted to compare the mean responses from older adults with mean responses from younger adults. All statistical analyses were carried out using Minitab Version 13.31 statistical analysis software.

Valero-Cuevas et al. (2003) developed a method to evaluate the S-D test used for quantifying the dynamic interaction between fingertip force magnitude (strength) and directional control (dexterity) during a pinch against a pinch meter. The metrics identified were pinch force magnitude (strength) and direction (dexterity). The method was based on the ability of participants to use pinch to fully compress a compression spring prone to buckling. A sufficiently slender compression spring will buckle when shortened below a critical length. The S-D test consists of asking participants to use key and opposition pinch to attempt to fully compress a set of springs with plastic end caps embodying a wide range of combinations of strength and dexterity requirements. After the participant attempts to compress each spring three times to its solid length using key or opposition pinch, a binary score is used to record if they succeeded at least once. The pinch force necessary to compress the spring to solid length defines the strength requirement. The ability to compress the spring without buckling defines the dexterity requirement. The subject pool consisted of 42 participants: 18 unimpaired adults under the age of 40 yr, 10 unimpaired adults over the age of 40 yr and 14 adults with carpo-metacarpal osteoarthritis (CMC OA) without neurological co-morbidities such as carpal tunnel syndrome. For both pinch styles, the S-D scores inside the core region were significantly higher for the older adults than for the CMC OA participants. Pinch meter readings were not significantly different across subject groups for either pinch style. S-D score was at least 94% reproducible. Moreover, the S-D test distinguished between CMC OA participants and asymptomatic older adults, while maximal pinch strength from pinch meter readings did not.

Beebe et al. (1998) presented a silicon-based force sensor packaged in a flexible package and described the sensors performance on human subjects. The metrics identified were Finger and hand force, pinch force for grasping activities. Silicon tactile sensors were used for measurement. The sensing element consisted of a circular silicon diaphragm over a sealed cavity with a solid Torlon dome providing force-to-pressure transduction to the diaphragm. The sensor design is based on a silicon diaphragm structure instrumented with ion-implanted piezo-resistors in a Wheatstone bridge configuration. The applied force is distributed across the diaphragm via the solid dome. The distributed force deforms the diaphragm giving rise to an output voltage proportional to the applied force for small deflections. At high forces, the diaphragm deflection is restricted by the cavity bottom limiting the maximum stress and extending the useful range of the sensor.

Dipietro et al. (2003) investigated the feasibility of using the Humanware Humanglove, a 20-position sensors glove, to measure finger’s range of motion (ROM), with particular regard to measurement repeatability. The metrics identified were fingers’ ROM for measuring repeatability. The Humanglove is a sensorized elastic fabric glove designed and commercialized by Humanware and is used for measuring the finger ROM. The Humanglove is equipped with 20 Hall Effect sensors that are distributed. Each sensor measures data related to a DOF of the hand. Four tests were conducted to measure repeatability - Mold Grip and Glove on Between Data Acquisition, Mold Grip and Glove off between Data Acquisition, Hand Flat and Glove on between Data Acquisition, Hand Flat and Glove off between Data Acquisition. The repeatability of measurements taken from the Humanglove is adequate to recommend the system for several applications in the field of rehabilitation engineering. The Humanglove can function as goniometric device for digit ROM acquisition. Moreover, as an additional advantage, the glove dynamically acquires data simultaneously from 20 hand degree of freedoms (DOF), including abduction and adduction (ABD/ADD) of fingers and thumb.

Welsh et al. (2001) evaluated the space suit glove to be used in and extravehicular activity (EVA) by using the standardized dexterity tests to provide objective measures of glove performance. The objective was to determine the effects of gloves on hand performance, to collect and examine range of motion data using an experimental data glove and to establish baseline data that will be used in future research. The metrics identified were performance time, range of motion, dexterity, strength, fatigue, and comfort. Hand performance data was collected for barehanded, unpressurized, and pressurized glove conditions for each standardized dexterity test. Hand tool test, Minnesota dexterity test and Purdue pegboard test were used to evaluate the biomechanics sensor data glove. The Statistical Analysis Software (SAS) package was used for the data analysis. The data glove was found to be an economical means of accurately measuring joint angles during simulated EVA tasks.

Figure 4: (a) Humanglove and (b) Biomechanics Sensor Glove

Kamieniarz et al. (1999) developed a computer program for the purpose of objectivisation of children hand dexterity. The program is composed of 6 subtests checking the control of the upper limb joints and the fine finger dexterity needed for mouse control. The metrics identified were control of the upper limb joints and the fine finger dexterity. Six tasks were developed to check manipulation dexterity of the hand - called blocks, labyrinth, ball, circle, board and centres. The program was written in MS Visual Basic 5.0 and operated on PC platform in Windows 95 system. The user interface is worked out in Polish. An analysis was made to check if the results differ significantly for boys and girls and the difference between computer- familiar children and computer-unfamiliar children. ANOVA and t-Student test were used for the statistical data analysis.

Small paragrhap describing the next topic or chapter

Reason for the project and that reson will be the hypothesis

Follow Dr. kondraske capability

The then you are proposing besides his ideas research something else such as basic experimentation plus the use of some semnsors BEP1, collect data using the sensors, then a model will be developed to make decisions. Helmet ar

Fix dexteriry to avoid having future problems on line or real time

Description of sensors and their use

METHODOLOGY

Pilot Study:

To identify the metric we conducted a pilot study on eight operators (working on A19/A1 manufacturing cell) and four managers at Johnson Controls, Juarez, Mexico. The main tasks during the pilot study were - video taping of the assembly station, the leveling & calibration station and the packaging station, the operator survey, the manager survey and measuring twice (before lunch and before end of the shift) the concentration, memory and reaction time of eight operators through simple computer games.

Equipment and Software: Digital camcorders, memory, reaction time, concentration software games.

Experiment Design: We used three digital camcorders to record the activities at the assembly station, the leveling & calibration station and the packaging station. These videos were transferred to DVDs (with the use of iMovie and iDVD software on the Macintosh OS) to examine and identify the metric. From the videos, “manual dexterity” was identified as the metric to be measured during the main study. Eight operators were asked to take a survey (Appendix A) which had questions about their daily tasks and discomforts in Spanish. Four managers were asked to take a survey (Appendix B) which had questions about decision-making policies, operator trainings, operator rotations and scheduling in English. The reaction time game consisted of a click of a mouse as soon as the operators see an object on the computer screen. This game was a direct indicator of the eye-hand coordination as it measured the time (in milliseconds) between a visual stimulus (object on the screen) and a response (mouse click). The memory game consisted of counting the moves the operators took to arrange 8 scrambled blocks in a 3x3 matrix form. This game was a direct indicator of problem solving capability in less number of trials and it was found to be a difficult test for most of the operators. The concentration game consisted of correctly guessing 2 same pairs out of 5 covered pairs of cards in less number of guesses. This game was a direct indicator of the short term memory and remembering capacity. This was also a difficult game for most of the operators. It was found that the operators do not really have to work fast and think for their tasks, but they have to have eye-hand coordination, two-arm coordination, hand-tool dexterity and assembly sequence remembering capacity. So dexterity and concentration were the metrics identified. But dexterity dominates concentration in case of the assembly tasks because there is more requirement of dexterously assembling the parts than remembering sequences which don’t have to be remembered after a few days assembly experience.

Data Analysis: From the reaction time plot shown in figure 6 (a), it can be observed that the reaction times before lunch were lower and by the end of the shift they were higher. It is good to have a lower reaction time for an operator. The workers work fast in the morning and work slow by the end of the day as they are tired.

Figure 6: (a) Average reaction Time Plot and (b) Concentration plot

From the concentration plot shown in figure 6 (b), it can be observed that more the experience of a worker in the facility more is his/her concentration except for one outlier (operator 3). Workers who are working for more number of years in the facility, guessed the same card pairs in less number of guesses while those with less number of years in the facility, guessed the same card pairs in more number of guesses. For less experienced workers, the concentration was more before lunch and less at the end of the shift while for more experienced workers, the concentration was almost same throughout the day.

From the operator’s survey, it is understood that the workers did not know the technical terms for the operations they were doing, the machines on which they were working and the sub-assemblies or parts they were using. All the workers like rotation, and going from one cell to another as they all like change in work. They all were uncomfortable in their standing positions as they stand all the day from 7.00 am to 4.30 pm and so they all need chairs. They feel that they do not have enough breaks (4 breaks a day) and they get more tired on Mondays and Fridays and mostly during the evenings. When a worker from one cell was relocated to another cell, one of the remaining workers had to cover up the work of the relocated worker. The worker, who covered up the relocated worker’s work, always worked on his/her previous work and his/her own work. They think that music and sitting can improve their productivity and can reduce their boredom. They have to work fast, remember sequences of assembling parts and do not have to reach distant parts while assembling. All the experienced workers (more than 2-3 years) have had injuries and the main body part affected was the wrist. All the workers are trained for 6 months and they give exams for each level gradually.

From the manager’s survey, it is understood that dexterity, concentration and reaction time were the most important metrics for human performance measurement. The managers generally do not look for who is doing mistakes in the cell. They look at the outgoing quality audit, scrap, first article sheet or the EOL audits to verify the non-defective products. They rotate/relocate workers based on the worker’s experience, certifications, ability to demonstrate a particular operation and observations. They do not use any software for their decision-making. The managers arrange monthly trainings, safety trainings, offline meetings, loan workshops and exams for certification in the facility to improve the worker’s performance and productivity.

Simulation:

The assembly station at Johnson’s Control was simulated at Human Performance Laboratory at New Mexico State University. A work table resembling a workbench was used for the assembly task and the dexterity tests. Assembly (according to the steps followed and parts used at Johnson Controls) of thermostats was conducted on the simulated work table. This simulated environment will enable us in future to conduct tests on the lab instead of at Johnson Control’s facility. The main tasks during the simulation were - assembly task and the dexterity tests. After every 20 real assemblies, the Purdue pegboard and hand too dexterity tests were applied to see if practice would improve the fine and gross dexterity.

Equipment: Purdue Pegboard for assessing fine dexterity (Model No.32020) and Hand-Tool Dexterity Test equipment (Model No.:32521) for assessing gross dexterity equipment from Lafayette Instruments as shown in figure 7 were used.

Figure 7: (a) Purdue Pegboard and (b) Hand Tool Dexterity Test Equipment

Experiment Design: The Purdue Pegboard test consisted of 4 tests - Right hand, left hand, both hands and the assembly tests. The right hand test consisted of inserting pins from the cup into the holes on the pegboard by using the right hand and counting the number of pins inserted in 20 seconds and number of pins dropped. The left hand test consisted of inserting pins from the cup into the holes on the pegboard by using the left hand and counting the number of pins inserted in 20 seconds and number of pins dropped. The both hands test consisted of inserting pins from the cup into the holes on the pegboard by using both hands and counting the number of pin pairs inserted in 20 seconds and number of pins dropped. The assembly test consisted of inserting a pin from the cup into a hole with the right hand and simultaneously picking a washer from the cup with the left hand, then putting the washer on top of the pin on the pegboard with the left hand and simultaneously picking a bolt from a cup with the right hand, then placing the bolt over the washer with the right hand and simultaneously picking another washer with the left hand and placing it on the bolt and then counting the number of assemblies made in 50 seconds, number of missed or wrong sequences and number of pins dropped. The hand-tool dexterity test consisted of unscrewing the screws on one side of the equipment and screwing it onto the other side of the equipment.

Data Analysis: From the Purdue pegboard test, it can be observed that as more number of real assemblies were made, the finger movement became efficient and faster (as more pins were found to be inserted into the holes), the range of bend of the fingers improved (as pick and place became easier), and lesser mistakes were made (as less pins were being dropped). From the hand-tool dexterity test, it can be observed that with more number of dexterous tasks done, speed of performing that task is increased and tools can be used comfortably.

During the simulation, the student (doing the assembly task of the worker) faced some difficulties initially like - distraction by background noise, pain in the legs/ankle, and then after some time faced strain on the eyes, pain in the back and shoulder. Later she could concentrate easily and started counting parts frequently to reach the pre-decided goal of making 20 assemblies. Thus, she had some dexterity capability but she kept assembling faster to reach the assembly task requirement, pre-decided by the managers.

Main Study:

Equipment and Software: Purdue Pegboard, Hand Tool Dexterity test equipment

Experiment Design:

Data Analysis:

RESULTS

DISCUSSION

As expected, the dexterity data from simulated and real environment for assembly station should not be significantly different using t - student test. Thus we are evaluating or assessing the simulated assembly station for dexterity. We may use the simulated assembly station for all other performance metrics that will be identified in future.

CONCLUSION

· Test equipment give an accurate but offline measurements for fine dexterity

· Future work on dexterity can be carried out in the laboratory => simulated environment is evaluated and is not significantly different from the real environment in terms of dexterity

· Depending upon the dexterity data, the managers can easily decide as to which operator is suitable for an assembly task

APPENDICES

A Operator Consent Form:

Operator Survey Questionnaire

February 18, 2005 at Johnson Controls, Juarez, Mexico

This survey will take about 10 minutes. Please circle the appropriate answer

1. What are your activities?

2. Do you do the same activities everyday? Yes No

3. How many years of experience do you have? 1-3 4-6 7-9

4. At what time do you start and end working - shift?

5. Do you get tired mentally (M) or physically (P)? M P

6. Which activities make you more tired?

7. What time of the day do you feel maximum tired? Morning Afternoon Evening

8. On which days do you put more efforts to work? Mon Tue Wed Thu Fri Sat

9. On which day do you feel more tired and why? Mon Tue Wed Thu Fri Sat

10. How many breaks do you take during the day?

11. Do you feel that you have enough breaks? Yes No

12. How do you find the work environment? (select as many as applicable)

Nice boring bossy relaxing stressful mention another

13. How often do you feel under stressed? everyday or just sometimes

14. What can make your work easier?

15. When you feel tired, what can be done to get you back to work?

16. Do you think work here is safe? Yes No

17. Do you get bored of your activities? Yes No

18. Did you get any training after you joined this company? Yes No

19. Does lighting or sound make you tired? Yes No

20. Do you have to reach far things and can you reach them easily? Yes No

Yes No

21. Do you have to remember many things Yes No

and do you remember them easily? Yes No

22. Do you have to work fast and can you work fast? Yes No

Yes No

23. How often do you forget the sequence of your activity?

Too often sometimes almost never

24. Do you get injured? Yes No

25. How often? Too often sometimes almost never

26. Is your standing/sitting position comfortable? Yes No

27. Do you change your position very often during work? Yes No

28. Can you concentrate on your activities? Yes No

B Manager Consent Form:

Cell Manager or Team Leader Survey Questionnaire

February 18, 2005 at Johnson Controls, Juarez, Mexico

This survey will take about 10 minutes. Please circle the appropriate answer

1. What are your activities in the company?

2. Do you do same activities everyday? Yes No

3. How many years of experience do you have as a manager? 1-3 4-6 7-9

4. At what time do you start and end working - shift?

5. Do you get tired mentally (M) or physically (P)? M P

6. Which activities make you more tired?

7. At what time of the day are you most tired? Morning Afternoon Evening

8. Do you schedule trainings for the workers and if yes, how? Yes No

9. How do you verify that the product is non-defective?

10. Can you find out easily who does mistakes? Yes No

11. How do you find who did the mistake if a product is defective?

12. What action do you take when you see that the workers are getting tired or bored?

13. What skills do you look for at each station? Which metric is useful at which station?

14. Do workers get stressed mentally (M) or physically (P)? M P

15. Rate the following metrics according to their importance for workers in your facility

· Memory

· Reaction time

· Concentration

· Attention

· Dexterity

· Range of motion or reachability

· Posture

16. Do you change the worker’s position when they are tired, Yes No

why, where and how often?

17. How do you decide as to which worker should work at which station?

18. What techniques or strategies do you use to measure the worker’s performance?

19. Do you use any software for the relocation of workers? Yes No

REFERENCES

Abdel-Malek, K., Yang, J., Yu, W., Duncan, J. (2000). Human Performance Measures: Mathematics. http://www.engineering.uiowa.edu/~amalek/papers/humanperformanceSAE.pdf

Abdel-Malek, K., Yu, W., and Duncan, J. (2005). Human Placement for Maximum Dexterity. SAE Digital Human Modeling and Simulation, June 14 - 16, 2005.

Beebe, D. J., Denton, D. D., Radwin, R. G., and Webster, J. G. (Feb, 1998). A Silicon-Based Tactile Sensor for Finger-Mounted Applications. IEEE Transactions on Biomedical Engineering, Vol. 45, No. 2, p. 151 - 159.

Dipietro, L., Sabatini, A. M., Dario, P. (April, 2003). Evaluation of an instrumented glove for hand-movement acquisition. Journal of Rehabilitation Research and Development, Vol. 40, No. 2, p. 179 - 190.

Kamieniarz, M., Stryla, W., Haglauer, P., Kamieniarz, G. (1999). Standardized Computer Tests for Assessment of Children Manual Dexterity. Computational Methods in Science and Technology, Vol. 5, p. 25 - 38.

Kondraske, G. V. (1995a). An elemental resource for human-task interface, International Journal of Technology Assessment in Health Care, Vol. 11, No. 2, p. 153 - 173.

Kondraske, G. V. (1995b). Human performance engineering: Section overview. In J. Bronzino (Ed.), The Biomedical Engineering Handbook, Boca Raton: CRC Press, pp. 2143 - 2145

Kondraske, G. V. (1995c). A working model for human system-task interfaces. In J. Bronzino (Ed.), The Biomedical Engineering Handbook. Boca Raton: CRC Press, pp. 2147 - 2164.

Kondraske, G. V., and Khoury, G. J. (1992). Telerobotic system performance measurement: Motivation and methods, SPIE Cooperative Intelligent Robotics in Space III, Bellingham, WA: SPIE, pp. 161-172.

Mital, A, Nicholson, A. S., Ayoub, M. M. (1993). A Guide to Manual material Handling, Taylor & Francis, London, Washington DC.

Pennathur, A., Contreras, L. R., Arcaute, K., and Dowling, W. (2003). Manual dexterity of older Mexican American adults: a cross-sectional pilot experimental investigation, International Journal of Industrial Ergonomics, Vol. 32, p. 419 - 431.

Pictures of Purdue pegboard and hand tool test dexterity equipment form Lafayette Instruments http://www.lafayetteinstrument.com/evaldexterity.htm.

Population Leadership Program, Project of Public Health Institute, office of Population.

Valero-Cuevas, F. J., Smaby, N., Venkadesan, M., Peterson, M., Wright, T. (2003). The strength-dexterity test as a measure of dynamic pinch performance. Journal of Biomechanics, Vol. 36, p. 265 - 270.

Welsh, M. H., and Akin, D. L. (2001). The Effects of Extravehicular Activity Gloves on Human Hand Performance, 2001 Society of Automotive Engineers, Inc.