A computer-assisted system for physical ergonomics analysis

A computer-assisted system for physical ergonomics analysis

Computers ind. Engng Vol. 20, No. 2, pp. 261-269, 1991 0360-8352/91 $3.00+ 0.00 Copyright © 1991 Pergamon Press pie Printed in Great Britain. All ri...

656KB Sizes 0 Downloads 4 Views

Computers ind. Engng Vol. 20, No. 2, pp. 261-269, 1991

0360-8352/91 $3.00+ 0.00 Copyright © 1991 Pergamon Press pie

Printed in Great Britain. All rights reserved

A COMPUTER-ASSISTED SYSTEM FOR ERGONOMICS ANALYSIS

PHYSICAL

JEN-GWO CHEN 1, ROBERT E. SCHLEGEL2 a n d J. BRIAN PEACOCK3 ~Department of Industrial Engineering, The University of Houston, Houston, TX 77004, 2School of Industrial Engineering, The University of Oklahoma, Norman, OK 73019 and 3CPC Group, Engineering North, General Motors Corporation, Pontiac, MI 48058, U.S.A.

(Received for publication 3 August 1990) Abstraet--A major obstacle to the widespread application of ergonomics knowledge to the working population is the inability to impart this knowledge to the ergonomics practitioner. Therefore, industry fails to derive maximum benefits from ergonomics research. The application of computer-assisted systems in ergonomics is one feasible solution to overcome this obstacle. This paper describes the development of an interactive computer-assisted Ergonomics Analysis SYstem (EASY). EASY is written in Foxbase Plus and QuickBASIC for IBM-PC compatible microcomputers. The system consists of three major components: the Ergonomics Information Analysis System (EIAS) for evaluation of tasks by the worker, the Physical Work Stress Index (PWSI) used by the supervisor or the ergonomist for further investigation of problem situations, and the Dynamic Lifting Analysis System (DLAS) for manual material handling tasks. Extensive use of menus for database entry/editing and analysis provides an efficient and friendly interface design. The system was evaluated by comparing the results of EASY and individuals with an introductory knowledge of ergonomics with experts' conclusions for nine test jobs involving a variety of physical work stressors. The evaluation indicated that 83% of EASY's disgnoses were accepted by the experts with some variation between individual experts and between EASY and the other diagnosticians.

1. INTRODUCTION

A major obstacle to the widespread application of ergonomics knowledge to the working population is the inability to impart this knowledge to the ergonomics practitioner. In advanced companies, members of the work force and supervisors have been given an introduction to the concept of ergonomics as it applies to their particular industry. This provides many individuals with the ability to recognize potential ergonomic problems. Solving the problems is a different matter as it requires the application of scientific data sources in a systematic fashion. This is more difficult to achieve for the individual with only an introductory knowledge of ergonomics. Shulman et al. [1] divided real world problem solving into four stages: problem sensing, problem formulating, searching, and problem solving. In ergonomics practice, as in many other areas, there is a large gap between the searching and problem solving stages. The stage model in Fig. 1 outlines an approach to fill this gap and defines the stages for physical ergonomics problem solving. Whatever the problem solving technique used, performance aids are needed to improve problem solving efficiency, particularly for inexperienced diagnosticians. In recent years, there has been substantial interest in the application of computer-assisted systems (e.g. expert systems, decision support systems) to support problem-solving and decision-making activities. In physical ergonomics analyses, the computer-assisted system can be used in a 'consulting' mode, in which solutions are generated based on various production rules and a question/answer dialogue between the user and the system. The programming languages used for computer-assisted system applications are generally either problem-oriented languages, such as FORTRAN, BASIC and PASCAL, or symbol-manipulation languages, such as LISP and PROLOG. Since Foxbase Plus, A dBASE compatible language with better performance than the original dBASE program, is a common commercial database management package with an internal structural language and control mechanism, it has been utilized in developing knowledge-based systems [2]. In the system presented here, Foxbase Plus is interfaced with QuickBASIC to provide control and a graphical output display. 261

262

JEN-Gwo CrlEN et al.

Job

Question/ T J Action measurement Reference / ~

data

|Ergonomist =~ vl J Ret rieva~% ;A J'~Retrieva k~ ; ; ConsuLtation

Memory

f Compu~erl assisted system

Fig. 1. Stage model of ergonomics problem solving. 2. COMPUTER-ASSISTED SYSTEMS IN ERGONOMIC ANALYSIS

Several knowledge-based systems have been developed in the analysis and design of manual materials handling tasks. Karwowski et aL [3-5] developed an expert system entitled LIFTAN to analyze the manual lifting tasks. LIFTAN includes three major modules: (a) a knowledge base, (b) an inference mechanism, and (c) a human interface. The knowledge module consists of task-related risk analysis (34 rules), operator-related risk (51 rules), task-related reasoning process and job redesign recommendation (30 rules), and operator-related explanation (34 rules). The domain knowledge were derived from the articles in the areas of epidemiology, anthropometry, biomechanics, work physiology, and psychologics of manual lifting. Kabuka et al. [6, 7] developed an expert system for the design of new and existing repetitive manual materials handling (RMMH) tasks. The maximum acceptable weight to be handled under a variety of worker and task variables are recommended for the new jobs. Possible solutions to the design of RMMH tasks are provided for unacceptable tasks. The knowledge base includes five types of manual materials handling tasks: lifting, lowering, carrying, pushing and pulling. Three major types of variables have been considered in this system. These are: (1) worker variables--age, sex, body weight, training and handling technique, (2) task variables--frequency, height, box length in the sagital plane, box width in the frontal plane, body twist angle for lifting and lowering tasks; frequency, height and horizontal travel distance for carrying, pushing, and pulling tasks, (3) environmental variables--temperature, humidity, air velocity and vibration. Laurig and Rombach [8, 9] developed an expert system entitled ERGON-EXPERT to detect and minimize health risks in manual materials handling tasks. After receiving investigated case facts, ergonomic analysis via five major ergonomic assessment are performed. These assessments are: (a) legal assessment (e.g. NIOSH's work practices guide for manual lifting), (b) biomechanical assessment to determine the torques and forces acting at the lumbo-sacral joint, (c) energetic assessment by analyzing metabolic load 'energy', (d) psychological assessment, and (e) psychophysical assessment. The redesign requirements and recommendations will be obtained based on the results from these assessments. Kengskool et al. [10] developed an expert system to help a user document a current system with a valid model and workplace diagram based on the work system information provided by the user. The system simulates several 3-D workplace diagrams. A summary of the results of system analysis related to area location, motion analysis, and urea utilization are provided. 3. ERGONOMICS ANALYSIS SYSTEM (EASY) CONFIGURATION

As with most computer-assisted systems, EASY organizes its structure into three levels: the data level, the domain knowledge level and the control level. An overview of the control structure is given in Fig. 2. There are three major subsystems. The Ergonomics Information Analysis System (EIAS) collects worker, task, and environment data from the worker's viewpoint [11]. The Physical Work Stress Index (PWSI) is used as a diagnostic tool by the supervisor or the ergonomist [12]. Potential problems and suggestions for improvement are identified by these two systems. If further investigation is needed and the job involves manual material handling, the Dynamic Lifting

A computer-assisted system for physical ergonomics analysis

Link from mdividu~ job ~ y ~ all~idm~ or model in ~e s y s ~ n

Control Level Knowledge Level peru Level

FoxbasePlus Plus I

Algorithm,produceonrulesor design guidelines for job analysis Interface with control level data storage anti ~ Database storage and reuieval

PWSI

~.BASI¢

EIAS and DLAS

Plus I

Main l~rognunto conu'ol function between model and data

Foxbase

Foxbase

263

Fig. 2. Overview of EASY's layered architecture.

Analysis System (DLAS) is used. The dynamic lifting model presented by Chen and Peacock [13] was adapted at the domain knowledge level for this subsystem. An outline of the basic structure for EASY's knowledge base is provided in Fig. 3. EASY is written in the forward chaining form. This means that the problem is investigated from initial conditions which are provided at the database level. Conclusions and recommendations are extended through several interactive stages. The knowledge modules of EASY are rule-based and consist of 45 production rules. Production rules are actually condition-action pairs. The system evaluates the conditions with reference to the created databases, and executes the action if the conditions are satisfied. The production rules in EASY are distributed among the three major subsystems, the EIAS (21 rules), the PWSI (20 rules) and the DLAS (4 rules). These three subsystems are independent. Thus, they may be used individually or in combination depending on the user and the purpose of the analysis. The general design guidelines of the EIAS help the user to understand basic ergonomic principles in job design. The PWSI profile indicates problems of high static loads and lack of postural variety as well as problems of excessive dynamic loads. Biomechanical analyses such as joint moments and segment linear velocity and acceleration are included in the DLAS as a part of the knowledge base.

3.1. Ergonomics Information Analysis System (EIAS) description The EIAS includes four sections: case identification, problem description and operatoroperation interaction. The case identification section records general personal and anthropometric data such as age, gender, height, and weight as well as job location. The problem description section records the incidence of injury resulting from the task, poor product quality, and/or poor quality of work life. The third section, job description, requests data describing the job, environment, workplace, parts and tools. The last section, operator-operation interaction, relates to problems EIAS User Self-Evalualion 1. Case Identification 2. Problem Description 3. Job Description 4. OperatorAon Interaction

PWSI Supervisor Investigation 1. Location 2. Orientation 3. Posture Base 4. Hand Position 5. Vibration

DI.AS For MamualMaterial Handling Task Analysis

1.LiR CapacityAnalysis 2. Biomechanical Analysis 3. NIOSH Crite~'iaAnalysis Fig. 3. Basic structure for EASY knowledge base. CAIE 20/2.--G

264

JEN-GwoCr~N et al.

of postural discomfort and includes biochemical, physiological, and information processing load descriptions of the job. The last two sections record quantitative data as opposed to the qualitative data collected in the first two sections. The quantitative data consists of a 5-point scale which describes the seriousness of each aspect of the problem. The theoretical basis for this self-evaluation checklist is the 'rang-theory' proposed by Borg [14]. In brief, this theory states that 'all individuals perceive roughly the same degree of exertion while performing dynamic work at their respective maximal physical capacity.' Ljungren [15] stated that the category-ratio is a reliable tool for making self-ratings. The following 5-point scale is used in the EIAS: Very Good 1

Good 2

Moderate

Bad

Very Bad

3

4

5

Very Good or Good means that there is no problem with the evaluated item. If the recorded rating is above the safety threshold (a scale value of 3), investigation or attention is necessary. Possible causes, suggestions, and comments are provided by the system through interactive dialogue and the production rules. The production rules for the EIAS are based on a checklist to investigate potential problems from the worker's self-evaluation. As an example, part of Rule 16 is given below: Rule 16--to identify possible causes of postural discomfort IF (1) Rating of postural discomfort exceeds safety threshold IF (1) Postural unit is standing base of support, and (2) Work dimension is restricted THEN (a) Standing too long in one place (b) Redesign workplace or moderately increase movement ELSE (a) Standing too long in one place (b) Provide appropriate seating IF (1) Posture unit is leg, and (2) Seating is not provided THEN (a) Standing too long in one place (b) Provide appropriate footrest ELSE (a) Sitting too long in one place (b) Provide improved seating IF (1) Posture unit is trunk THEN (a) Trunk curved forward when sitting or standing (b) Keep trunk upright or move task objects closer to body. etc. This rule traces through various posture units (base, leg, trunk, shoulder, neck, arm, wrist, and hand) and provides possible causes and suggestions where appropriate. The general design guidelines related to the investigation item are provided concurrently to teach the user how to avoid unnecessary problems and improve performance. The design principles forming the basis for Rule 16 follow: 1. 2. 3. 4. 5. 6.

Avoid any kind of bent or unnatural posture Avoid keeping the arms outstretched Provid an optional, seated workplace Use supports under elbows, forearms or hands to avoid fatigue Maintain a balance between arm movements on both sides of the body Use back and leg rests for sedentary work.

3.2. Physical Work Stress Index ( P W S I ) description The PWSI is an observational method of physical work stress analysis which possesses the ease of application of traditional work study techniques, but also provides better accounting of human and task variables. The PWSI method involves the following components: (1) movement required to accomplish a job, (2) orientation with regard to the primary work place, (3) posture base,

265

A computer-assisted system for physical ergonomics analysis Table 1. Classification of physical work stress using PWSI Category PWSI

Table 2. Critical load factors for PWSI components Instantaneous Static/dynamic PWSI component pload load

Very low Low Moderate High Very high Extremely high

Movement Orientation Postural base Left hand position Right hand position Acceleration Temperature

0-2 2-4 4-6 6-8 8-10 > 10

1.080 0.432 1.008 1.800 1.800 0.432 0.648

0.315 0.126 0.294 0.525 0.525 0.126 0.189

(4) hand position, (5) acceleration (vibration), (6) thermal load and (7) external load. The technique includes the process of activity sampling of the above physical work components. The static/ dynamic load factor describes the change, or lack of change, in each of the components of physical work stress. The instantaneous load factor describes the physical work stress averaged over a series of snapshots of the task. The static/dynamic load, instantaneous load, and PWSI values are derived from these data. The production rules for the PWSI are classified into three categories based on a regression model of the PWSI as a function of increases in heart rate for various levels of physical stress [16]. Rules 1-6 are used to categorize physical work load based on the PWSI value (Table 1). Rules 7-13 are applied to identify possible causes for instantaneous loading (Table 2). Rules 14-20 are used to identify possible causes of excessive static or dynamic loads (Table 2). Rule 17 is given below as an example. Rule 17 IF (1) Value of PWSI is equal to or greater than 4.29, and (2) Value of static/dynamic load for left hand position is equal to or greater than 0.525 THEN (a) Some operations are out of reach of the left hand (b) Redesign workplace to arrange materials and tools inside the reach area.

3.3. Dynamic Lifting Analysis System (DLAS) description The DLAS includes three components: lifting capacity analysis, biomechanical analysis and NIOSH guidelines analysis. For a detailed discussion of the system, refer to Chen and Peacock [13]. The principle of static equilibrium is used to determine lifting feasibility. Thus, the production rules for this system are based on this principle. The NOISH guidelines [17] which were developed following extensive epidemiological, biomechanical, physiological and psychological analysis, were used to define the three zones of lifting safety (acceptable lifting conditions, administrative controls required, and hazardous lifting conditions). Suggestions are provided following the analysis. 4. S O F T W A R E D E S I G N

EASY is written in Foxbase Plus and QuickBASIC. The basic software structure for EASY is shown in Fig. 4. The major control menu is EASY.PRG. Three major subsystems EASY.PRG Control Program for EASY 1. System Description 2. EIAS Analysis 3. PWSI Analysis 4. DLAS Analysis 5. Exit

I

EIAS.PRG Master File for EIAS I. Create EIAS Database 2. Edit EIAS Database 3. EIAS Analysis 4. Print Analysis Results 5. Exit EIAS to Main Menu

I

PWSI.PRG Master File for PWSI 1. C~ate New PWSI File 2. Print Existing File 3. Edit Existing File 4. PWSI Analysis 5. Polar Coordinate Display 6. Print Analysis Results 7. Exit PWSI to Main Mc~u

I

DLAS.PRG Master F'tle for DLAS 1. Lifting Capacity Analysis 2. Biomechanical Analysis 3. NIOSH Criteria Analysis 4. Exit DLAS to Main Menu

Fig. 4. Basic software structure for EASY.

266

JEN-GWo CrlEN et al.

I

CASE IDENrwtCATION DATA ENTRY SCREEN

I

Date (mm~/dd/yy)

11/12/86

Tm~ 0ah:mm)

10:25

Job Code (4 digits)

2413

Case Number (4 digits)

3524

IEER

Section (4 digits)

4823

Worker ID (4 digits) 6267 Age 30 Height (in) 67

Shift(I,2or 3) Gender (M or F) Weight (Ib)

I M 170

I~nt

(4 chars)

Use cm'sor keys to edit data. Press and at the same time to end data entry

Fig. 5. An example of input/edit in EASY.

follow: (1) EIAS.PRG for EIAS analysis, (2) PWSI.PRG for PWSI analysis, and (3) DLAS.EXE for biomechanical lifting analysis. Each subsystem has its own control menu to allow the user to enter, edit, retrieve, analyze and print data. Each subsystem also has independent database structures for each section. For example, four separate database structures are created in EIAS corresponding with the four sections mentioned previously but these databases can be connected if necessary. The input/edit screens (Fig. 5) are designed to offer the user a friendly data entry and editing environment. The system allows the user to retrieve any previous data file for updating or analysis. Interactive commands help the user to perform the entire process efficiently. The results are presented in tabular form (Fig. 6). A unique aspect of the PWSI software is the graphic display of results using a weighted polar coordinate display (Fig. 7). The weighted polar graph uses the sector angle to indicate the component weighting or importance and the radius length to represent the component value [18]. Therefore, both area and shape are used to differentiate normal and deviant conditions. A brief example of EASY is shown in the Appendix. 5. SYSTEM E V A L U A T I O N

Evaluation refers to the determination of the system's performance for the specific consultation task for which it was designed. The process of evaluation should also ensure that the end user will be satisfied with the system's performance. There is no point having a powerful computer-assisted system if the end user cannot effectively and efficiently interact with it. Therefore, interface and performance evaluations of a computer-assisted system are required to determine the system's ease of use and accuracy of diagnosis.

5.1. User interface evaluation A total of twenty subjects participated in an evaluation of the EASY user interface. Several subjects were experienced in human-computer interface design. Each subject was made familiar with EASY using the online system description and instructions from the experimenter. A series of 35 mm slides of a dishwashing task was shown to describe a sample job. Some of the task variables such as pan weight and case weight were also provided. The subjects were asked to study the task and to interact with the software based on their perception of the task. After using the system, the subjects filled in questionnaires relating their impressions about various interface issues.

I STATIC/DYNAMIC L O A D TABLE Job Code: Wctkex ID: Exmrnal Load: Location: L Hand Position: Postur© Base: Temperan~:

2413 6267 4.00 0.90 3.10 0.09 0.09

Orientation: R Hand Position: Vibration:

I

Thermal

Load

2.15 3.10 2.09

Overall Static/Dynamic Load: 1.96 Right hand position

Fig. 6. An example of output screen in EASY.

Lett

hand posit.ion

Fig. 7. Weighted polar coordinate display.

A computer-assisted system for physical ergonomics analysis

267

Table 3. Test jobs for performance evaluation Job

Type

1

MMH

Description

2

MMH

3

MMH

4

Secretarial

5

Hand tool

6

Hand tool

7

Light assembly

8

Dishwashing

9

Book binding

Lift 9 Kg tote box form low cart, walk 25 m, place on high cart; 15 cycles/hr; 22°C, 60% RH Lift 18 Kg tote box from low cart, walk 50 m, place on high cart; 10 cycles/hr, 32°C, 20% RH Lift 6 Kg box from upper cart, place on lower cart, push 45 Kg cart 15 m; 15 cyles/hr; 14°C, 50% RH Typing, filing, writing, consulting and word processing; 25°C, 50% RH Overhead operation of 1.4 Kg power screwdriver, 1 cycle/rain; 25°C, 40% RH Waist height operation of 1.4 Kg power screwdriver; l cycle/min; 25°C, 40% RH Assembly of fishing reels from parts in bins at workplace; 24°C, 50% RH Lifting 0.5--16 Kg), washing, pulling of dishes and pins; 30°C, 450 RH Removing old covers, making labels and covers, gluing, etc.; 25°C, 40% RH

All subjects approved of the formats used by the system and considered the system easy to use. One subject suggested that the output for hardcopy be modified to allow for a short or detailed summary. One subject felt that the conclusions of the system were not helpful.

5.2. System performance evaluation The technique of matching the conclusions of the computer-assisted system with those of the expert was applied for system performance evaluation. Since it is often difficult for the expert and the computer-assisted system to obtain identical unique results and suggestions for all situations, the experts were asked whether they agreed with EASY's results or not. The EASY performance evaluation involved two stages. In the first stage, seven diagnosticians (two industrial engineering students with an ergonomics background, two industrial engineering students without an ergonomics background, two engineering students without an ergonomics background and with the help of EASY) examined nine test jobs involving a variety of physical work stressors (Table 3). In the second stage, two evaluators (experts) assessed the diagnosis results without knowing the identity of the diagnosticians and the knowledge-based system. The nine jobs were reviewed by the diagnosticians using 35 mm slides. A description of task variables and ergonomics information from the worker's self-evaluation were also provided. The evaluation form for the experts consisted of two portions: the EIAS evaluation and the PWSI evaluation. The results from the diagnosticians, the job description, the ergonomics information and the slides for each job were presented to the evaluators. The diagnosis results were coded in a standardized format and placed in random order to disguise the identities of the individual diagnosticians. The 126 diagnoses were classified by each evaluator as acceptable or unacceptable [19]. Analysis of variance was used to analyze the overall differences between EASY and the other diagnosticians. Similar tests were used to demonstrate individual differences between the six human diagnosticians. Overall performance was computed as the average of the performance of the PWSI and the EIAS. Table 4 presents the evaluators' ratings for each diagnostician. Overall, the experts agreed with the results of the computer-assisted system in 83% of the cases (94% for EIAS and 72% for PWSI). Table 4. EASY performance evaluation results

Diagnostician

Number (%) of jobs deemed acceptable by experts (PWSI)

Number (%) of jobs deemed acceptable by experts (PWSI)

Number (%) of jobs deemed acceptable by experts (PWSI)

EASY 1 (IE/ERGO) 2 (IE/ERGO) 3 (IE) 4 (IE) 5 (ENGR) 6 (ENGR)

13 (72) 11 (61) 7 (39) 11 (61) I I (61) 7 (39) 10 (56)

17 (94) 12 (67) 9 (50) 10 (56) 10 (56) 12 (67) I0 (56)

30 (83) 23 (64) 16 (44) 21 (58) 21 (58) 19 (53) 20 (56)

268

JEN-Gwo CrlEN et al.

A Duncan multiple range test revealed a significant advantage of EASY over the other diagnosticians at 0t = 0.10. The ANOVA also determined that there was a significant difference between the experts' rating at ~t = 0.05. 6. CONCLUSIONS

The main purpose of this research was to develop a computer-assisted system with a knowledge base to allow a worker or supervisor with a minimal ergonomics background to diagnose and solve physical ergonomics problems. The resulting system can also be applied in workplace design and personnel selection and placement (e.g., using DLAS). EASY can be enhanced to support a specific application by restricting the type of tasks that are analyzed and incorporating knowledge about assembly processes pertinent to a particular industry. The knowledge base can be modified by providing additional sophisticated rules to handle subtle aspects of particular cases. This modification may involve either the production rules or the data file structure to make the system compatible with existing data files in a company. Life many professions, ergonomics has now reached a stage where its techniques are being applied by practitioners who have not had the benefit of substantial education and training in the field. Clearly, the danger lies in the mis-application of the techniques through inadequate analysis of problems and improper selection and implementation of solutions. It is believed that the EASY presented here is a positive step toward the widespread distribution of ergonomics design principles and improved working conditions. REFERENCES 1. L. S. Shulman, M. J. Loupe and R. M. Piper. Studies of the Inquiry Process: Inquiry Patterns of Students in Teacher Training Programs. East Lansing: Educational Publications Services, Michigan State University (1968). 2. M. Williamson. Artificial Intelligence for Microcomputers--The Guide for Business Decisionmakers. Brady, New York (1985). 3. W. Karwowski, T. L. Ward, L. E. Palenque and S. E. Wisman. Building an expert system for ergonomics application: A LIFTAN experience. In Trends in Ergonomics/Human Factors Ili (Edited by W. Karwowski). North-Holland, Amsterdam, pp. 51-61 (1986). 4. W. Karwowski, N. Mulholland, T. L. Ward, V. Jagannathan and R. L. Kirchner. LIFTAN: An experimental expert system for the analysis of manual lifting tasks. Ergonomics 29, 1213-1234 (1986). 5. W. Karwowski, A. Mital, L. E. Palenque and T. L. Ward. Development of a microcomputer-based expert system for the analysis of manual material handling tasks in industrial settings. Ind. J. Ind. Ergon. 2, 49-59 (1987). 6. A. M. Genaidy, S. S. Asfour and M. Kabuka. An expert system for the design of manual materials handling tasks. In Trends in Ergonomics~Human Factors IV (Edited by S. S. Asfour). North-Holland, Amsterdam, pp. 559-566 (1987). 7. M. Kabuka, A. M. Genaidy and S. S. Asfour. A knowledge-based system for the design of manual materials handling. Appl. Ergon. 19(2), 147-155 (1988). 8. V. Rombach and W. Laurig. ERGON-EXPERT--A knowledge-based approach to the design of workplaces. In Trends in Ergonomics~Human Factors V (Edited by F. Aghazadeh). North-Holland, Amsterdam, pp. 53-61 (1988). 9. W. Laurig and V. Rombach. Expert systems in ergonomics: requirements and an approach. 32, 795-811 (1989). 10. K. Kengskool, J. Goldman and M. S. Leonard. An expert system for human operator's workplace design. In Trends in Ergonomics~Human Factors IV (Edited by S. S. Asfour). North-Holland, Amsterdam, pp. 567-573 (1987). 11. J. Chen, J. B. Peacock and R. Schlegel. A database management system for ergonomics information analysis. In Trends in Ergonomics~Human Factors V (Edited by F. Aghazadeh). North-Holland, Amsterdam, pp. 3-14 (1988). 12. J. Chen, R. Schlegel and J. B. Peacock. An observation technique for physical work stress analysis. Int. J. Ind. Ergon. 3, 167-176 (1989). 13. J. Chen and J. B. Peacock. An interaction biomechanical lifting model. In Trends in Ergonomics/Human Factors H (Edited by R. E. Elbert and C. G. Elbert). Elsevier, Amsterdam, pp. 527-542 (1985). 14. G. Borg. A category scale with ratio properties for intermodal and inter-individual comparisons. In Psychophysical Judgement and the Process of Perception (Edited by H. Geissler and P. Petzold). VEB Deutscher Verlag der Wissenschaften, Berlin, pp. 25-34 (1982). 15. G. Ljungren. Observer ratings of perceived exertion in relation to self ratings and heart rate. Appl. Ergon. 17(2), 117-125 (1986). 16. B. J. Peacock and J. Chen. A physical work stress index, technical report. Department of Industrial Engineering, University of Oklahoma (1986). 17. National Institute for Occupational Safety and Health, Work Practices Guide for Manual Lifting. Washington, D.C., U.S. Government Printing Office, No. 81-122 (1981). 18. J. Chen, D. Deal and V. Jeyakumar. A comparison of three display formats for multi-dimensional systems. Trends in Ergonomics~Human Factors V (Edited by F. Aghazadeh). North-Holland, Amsterdam, pp. 101-108 (1988). 19. V. L. Yu, B. G. Buchanan, E. H. Shortliffe, S. M. Wraith, R. Davis, A. C. Scott and S. N. Cohen. An evaluation of the performance of a computer-based consultant. Computer Prog. Biomed. 9, 95-102 (1979).

A computer-assisted system for physical ergonomics analysis

269

APPENDIX Job description The worker is responsible for drilling, reaming and coldworking holes to close tolerances using the driller, riveting guns and other riveting devices in the forward fuselage section, cockpit and wing area. EIAS mode PROBLEM DESCRIPTION DATA ENTRY SHEET For Injury: injury injury injury injury

severity is very mild parts are hand and arm types are strain/sprain causes are others

For Product Quality: poor product quality is very mild poor quality causes are awkward workplace For Work Life Quality: poor work life quality is very mild poor quality types are complaint JOB DESCRIPTION DATA ENTRY SHEET For Job Organization: job complexity is very good job repetition rate is moderate social interaction is moderate job satisfaction is moderate For Job Workplace: reach is moderate fit is moderate obstacle is moderate For Parts: size is good weight is good material is good For Tool: handling is moderate vibration is moderate weight is good OPERATION/ION INTERACTION DATA ENTRY SHEET Posture Discomfort is level 4 Posture Discomfort Units are neck, arm, wrist, trunk Biomechanical Static Load is level 4 Biomechanical Dynamic Load is level 2 Sensory Load is level 2 Cognitive Load is level 2 Motor Load is level 2 EIAS ANALYSIS FOR JOB DESCRIPTION Factors above the safety threshold is noise at level 4 Can the noise level can be reduced (Y/N)? Y Possible Cause: workplace is too noisy Suggestion: control or isolate noise resource EIAS ANALYSIS FOR OPERATOR/ION INTERACTION Factors above the safety threshold are; posture discomfort units are neck, arm, wrist and trunk posture discomfort is level 4 static load is level 4 Possible Cause: head excessively inclined backwards or forwards Suggestion: avoid unnatural or head position Possible Cause: posture is unnatural Suggestion: keep posture upright or provide appropriate rests P W S I mode Overall Instantaneous Load: 3.41 Overall Static/dynamic Load: 0.46 Overall PWSI: 1.72 Comment: the physical work load for this job is very low and is acceptable right and left hand position contribute most of the stress