© 1985 Thomas Blair DeGreve 



A WORKPLACE DESIGN EXPERT SYSTEM 

by 

THOMAS BLAIR DeGREVE, B.S. in I.E. 

A THESIS 

IN 

INDUSTRIAL ENGINEERING 

Submitted to the Graduate Faculty 
of Texas Tech University in 
Partial Fulfillment of 
the Requirements for 

the Degree of 

MASTER OF SCIENCE 

IN 

INDUSTRIAL ENGINEERING 

Approved 

C h a i r m a n of tzhe C o m m i t t e e 

/^ ^̂ Ŵ̂  

LJJAJM-^UV^ Inn • h i OV''^^ 

Deaiy of the Graduate School 

December, 1985 



1 • 

/ f ^ ^ TABLE OF CONTENTS 

LIST OF TABLES iv 

LIST OF FIGURES v 

I. INTRODUCTION 1 

Scope 1 

Review of Previous Research 4 

Expert Systems 4 

M^orkplace Design 10 

II. SYSTEM CONCEPTUALIZATION 15 

System Components 15 

The Context Tree 19 

System Prototype 23 

III. KNOWLEDGE ENGINEERING 24 

Knowledge Acquisition 24 

Knowledge Representation 26 

IV. DEVELOPMENT OF THE SYSTEM 40 

The Development Engine 40 

Description of the Data 43 

Construction of the Knowledge Base . . . . 46 

The Inference Engine 47 

V. CONCLUSION 51 

LIST OF REFERENCES 54 

1 1 



APPENDIX 

A. DEFINITION OF KNOWLEDGE BASE PROPERTIES . . . . 56 

B. LISTING OF THE KNOWLEDGE BASE 59 

C. SAMPLE CONSULTATIONS 95 

D. DEFINITION OF ANTHROPOMETRIC MEASUREMENTS . . . 108 

1 1 1 



LIST OF TABLES 

Table 1 : An Overview of the Key Events in the History 
of Artificial Intelligence 

Table 2: Ten Generic Categories of Expert Systems . . . 11 

Table 3: Rules for Determining the Workstation Type . . 28 

Table 4: Corrected Anthropometric Measurements . . . . 45 

IV 



LIST OF FIGURES 

Figure 1 

Figure 2 

Figure 3 

The Context Tree 

The Architecture of an Expert System . . . 

20 

41 

Major Categories of Search Strategies used by 
Inference Engines 48 

Figure 4: Sketch of Anthropometric Measurements . . . . 109 



CHAPTER I 

INTRODUCTION 

Scope 

The field of ergonomics is rapidly becoming a key area 

of interest to employers who are concerned with providing a 

comfortable, safe, and pleasant working area for their 

employees as well as for themselves. The interest in apply-

ing ergonomic principles to industrial workplaces is most 

likely a result of recent correlations made between the 

design of a workplace and employee health or productivity 

(Tichauer 1975, Konz 1979, Grandjean 1982). However most 

companies do not possess the expertise necessary for apply-

ing ergonomic principles to the design of workplaces. Even 

with the abundance of reference material available, most 

facility engineers find it difficult to develop a proficien-

cy in defining the work classification, analyzing the task 

requirements, quantifying the work force population, and 

finally coming up with an ergonomically sound workplace 

design. While theoretical research needs to continue, a 

practical method of putting the present knowledge to use is 

also important. The intention of this research is to 

develop a computer program called an expert system which 

will provide practical proficiency or expertise for the 

workplace designer. 

Expert system technology is an application of a branch 

1 



of artificial intelligence research concerned with develop-

ing programs that use symbolic knowledge to simulate the 

behavior of human experts. Professor Edward Feigenbaum 

(1977) of Stanford University, one of the leading research-

ers in expert systems, has defined an expert system as: 

... an intelligent computer program that uses 
knowledge and inference procedures to solve prob-
lems that are difficult enough to require 
significant human expertise for their solution. 
Knowledge necessary to perform at such a level, 
plus the inference procedures used, can be thought 
of as a model of the expertise of the best practi-
tioners of the field. 

The knowledge of an expert system consists of 
facts and heuristics. The "facts" constitute a 
body of information that is widely shared, public-
ly available, and generally agreed upon by experts 
in a field. The "heuristics" are mostly private, 
little discussed rules of good judgment (rules of 
plausible reasoning, rules of good guessing) that 
characterize expert-level decision making in the 
field. The performance level of an expert system 
is primarily a function of the size and the quali-
ty of a knowledge base it possesses. 

Feigenbaum calls those who build knowledge-based expert 

systems "knowledge engineers" and refers to their technology 

as "knowledge engineering." Expert systems are knowledge-

intensive computer programs with the following dimensions: 

expertise, symbol manipulation, uncertainty, complexity, 

reasoning, and explanation. They use rules of thumb, or 

heuristics, to focus on .the key aspects of a particular 

problem, manipulate symbolic descriptions of the problem, 

and apply reasoning to the knowledge they have been given 

concerning the problem, to reach a conclusion. They often 



consider a number of competing hypotheses simultaneously, 

and frequently make tentative recommendations or assign 

weights to alternatives. 

Current expert systems are confined to well-

circumscribed tasks. They are not able to reason over a 

broad field of expertise. They cannot reason from axioms or 

general theories. They do not learn and, thus they are 

limited to using the specific facts and heuristics that they 

were "taught" by a human expert. They lack common sense, 

they cannot reason by analogy, and their performance deter-

iorates rapidly when problems extend beyond the narrow task 

they were designed to perform. On the other hand, knowledge 

systems do not display biased judgments, nor do they jump to 

conclusions. They always attend to details, and they always 

systematically consider all of the possible alternatives 

(Harmon and King 1985). 

It is reasonable, then, to consider developing an 

expert system to assist facility engineers in the design of 

industrial workplaces. Workplace design is a problem domain 

relatively narrow enough, yet unstructured and uncertain 

enough to merit development of a dedicated expert system. 

Workplace design literature offers many guidelines and prin-

ciples to the potential designer but there are no clear-cut 

procedures designed for direct application in an industrial 

setting, A workplace design expert system has the potential 

to utilize guidelines from the literature, borrow experience 



from a human expert, and apply this knowledge in a practical 

sense. 

The progress of this project was aided by reference 

material (Hertzberg 1972, Ayoub 1973, Konz 1979, Grandjean 

1982) and the advice of Professor M. M. Ayoub, an experi-

enced workplace design expert. 

Review of Previous Research 

In order to proceed with the development of a workplace 

design expert system, an extensive review of previous 

research in the areas of both expert systems and workplace 

design is appropriate. The following sections will summar-

ize previous research in these two areas, respectively. 

Expert Systems 

As World War II ended, separate groups of British and 

American scientists were working to develop what is now 

called a computer. Each group wanted to create an electro-

nic machine that could be guided by a stored program of 

directions and made to carry out complex numerical computa-

tions. The principal British scientist, Alan Turing, argued 

that such a general-purpose machine would have many differ-

ent uses. Turing argued that the fundamental instructions 

given to such a machine ought to be based on logical opera-

tors, such as "and," "or," and "not." These general 

operators could then be used to assemble the more special-



ized numerical operators needed for arithmetic calculations. 

Moreover, programs based on logical operators would be 

capable of manipulating any type of symbolic material that 

one might want to work with, including statements in ordi-

nary language. 

The more practical American scientists, knowing the 

machine was going to be expensive to build and assuming they 

would not build very many, decided against using logical 

operators. They were confident they were building a machine 

that would do only arithmetic calculations; therefore, they 

chose to use numerical operators, such as "+", "-", and ">". 

This decision, which the British subsequently followed as 

well, resulted in large computers that are essentially very 

fast calculating machines. 

Until very recently, the decision to use numerical 

operators seemed like a reasonable one to most people 

involved with computers. However, a small group of computer 

scientists continued to explore the ability of computers to 

manipulate non-numerical symbols. At the same time, psycho-

logists concerned with human problem solving sought to 

develop computer programs that would simulate human behav-

ior. Over the years, these people have formed the inter-

disciplinary subfield of computer science called artificial 

intelligence (AI). AI researchers are concerned with 

developing computer systems that produce results that would 

normally be associated with human intelligence (Harmon and 



King 1985). 

The applied side of artificial intelligence research 

includes the areas of natural language processing, robotics, 

and expert systems. The latter has enjoyed much success in 

recent years and initiated a rush to find practical applica-

tions for this new technology. Table 1 presents an overview 

of the key events in the history of artificial intelligence 

with particular attention focused on expert systems 

research, 

The earliest acknowledged expert system, DENDRAL, is a 

chemistry expert system designed to examine a spectroscopic 

analysis of an unknown molecule and predict the molecular 

structures that could account for that particular analysis. 

In 1964, Joshua Lederberg, a Nobel prize-winning chemist, 

developed the DENDRAL algorithm. In 1965, with the aid of 

Edward Feigenbaum and Bruce Buchanan, the DENDRAL expert 

system was programmed directly in the LISP computer language 

(Lindsay, Buchanan, Feigenbaum and Lederberg 1980). 

Probably the most well-known expert system, MYCIN, is a 

computer program designed to provide attending physicians 

with advice comparable to that which they would otherwise 

get from a consulting physician specializing in bacteremia 

and meningitis infections. MYCIN was the first large expert 

system to perform at the level of a human expert and to pro-

vide users with an explanation of its reasoning. MYCIN was 

developed at Stanford University in the mid-1970's (Buchanan 



Table 1. An Overview of the Key Event in the 
History of Artificial Intelligence (from 
Harmon and King 1985) 

P E R I O D 

Pre-World War II 

Pre-AI 

The formative years, 
1955-1960 
Initiation of AI 

The years of 
development, 
1961-1970 

Search for general 
problem solvers 

The years of 
specialization, 
1971-1980 

Discovery of 
knowledge based 
systems 

Rush to applications 
1981-

International 
competition and 
commercialization 

KEV E U E N T S 

Formal. Logic 
Cognitive Psychology 

Comjp.uters Developed 
N. Wiener, "Cybernetics" 
A.M. .Turing, ""Computing And 
Intelligence^ 

Information processing Lansuase I 
Dartmouth Seminar on AI,l95o 
General Problem Solver (GPS) 

LISP 
Heuristics 
Robotics 
Chess programs 
Dendral (Stanford) 

MYCIN (Stanford) 
Hearsay II (Carnegie-Mellon) 
MACSYMA (MIT) 
EMYCIN (Stanford) 
PROLOG 

PROSPECTOR (SRI) 
Japan's Fifth-Generation Project 
E.Fergenbaum, "The Fifth 
Generation" 

INTELLECT (AIC) 



8 

and Shortliffe 1984). 

A line of expert system development beginning with 

SAINT (Slagle 1961) culminated with MACSYMA. The design for 

MACSYMA was originally laid out in 1968 by Carol Engleman, 

William Martin, and Joel Moses at MIT and has been under 

continual development since 1969. MACSYMA performs differ-

ential and integral calculus symbolically and excels at 

simplifying symbolic expressions. It incorporates hundreds 

of rules, each of which expresses one way to transform an 

expression into an equivalent. The solution to any problem 

requires finding a chain of rules that transforms the origi-

nal expression into one that is suitably simplified (Martin 

and Fateman 1971). 

HEARSAY was developed to demonstrate the possibility of 

a speech-understanding system. Development began in the 

late 1960's at Carnegie-Mellon University. By the time the 

HEARSAY II project was finished in 1975, a system had been 

developed that could deal with a limited amount of spoken 

grammar and a vocabulary of about 1,000 words. HEARSAY II 

demonstrated the clear superiority of symbolic, heuristic 

methods over statistical methods in dealing with problems 

involving meaning (Erman et al. 1980), 

PROSPECTOR was developed in the late 1970's at Stanford 

Research Institute, International by a team" that included 

Peter Hart, Richard Duda, R. Reboh, K. Konolige, P. Barrett, 

and M. Einandi. The development of PROSPECTOR was funded by 



the U.S. Geological Survey and by the National Science Foun-

dation. PROSPECTOR was designed to provide consultation to 

geologists in the early stages of investigating a site for 

ore-grade deposits. The program informs users of possible 

data interpretations and identifies additional geological 

observations that would be valuable to reach a more definite 

conclusion (Duda and Reboh 1984). 

PUFF was built in 1979 using the EMYCIN knowledge sys-

tem building tool at Stanford University. PUFF was designed 

to interpret measurements from respiratory tests admini-

stered to patients in a pulmonary function laboratory. PUFF 

interprets a set of test results and produces a written 

statement that includes a set of interpretations and a diag-

nosis for the patient (Aikens, 1984). 

Recent systems built since 1980 are mostly commercial 

systems and include the following: XCON (RI) and XSEL— 

knowledge systems that help configure computer systems for 

clients, GENESIS--a package of systems that help molecular 

geneticists plan DNA experiments, DELTA/CATS--an expert sys-

tem that aids locomotive maintenance personnel, and DRILLING 

ADVISOR--an expert system that helps oil rig supervisors to 

solve problems. 

There are many tools available commercially to help 

build expert systems. Most tools provide an inferencing 

mechanism along with the means for the user to enter his 

knowledge base. Some tools provide help screens and debugg-



10 

ing utilities. A list of some of the current expert system 

development tools include: EXPERT, KES, 0PS5, S.l, TIMM, 

ART, KEE, LOOPS, ES/P ADVISOR, Expert Ease, INSIGHT, M.l, 

and Personal Consultant, 

Most expert systems fall into ten generic categories of 

knowledge engineering applications. These ten categories, 

presented in Table 2, are based on the types of problems 

that different expert systems address. The subject system 

of this thesis falls into the Design-type category. 

v̂  Workplace Design 

Interest in industrial workplace design can be traced 

back to the 1880's when Frederick Taylor and Frank and 

Lillian Gilbreth began their work with time and motion 

studies. The primary objective of their work involved mea-

suring and improving worker productivity. Consequently, 

they recognized the importance of workplace design, but they 

did not recognize its full importance. Das and Grady (1983) 

recognize the following five objectives in designing indus-

trial workplaces: 

/I. Measure and improve worker productivity; 

2. Enhance worker satisfaction and job attitudes; 

3. Reduce operator fatigue; 

4. Improve working environments; 

5. Minimize worker safety hazards.\ 



11 

TABLE 2: Ten Generic Categories of Expert Systems 
(from Hayes-Roth et al. 1983) 

CRTEGORV 

INTERPRETATION 

PREDICTION 

DIAGNOSES 

DESIGN 

PLANNING 

MONITORING 

DEBUGGING 

REPAIR 

INSTRUCTION 

CONTROL 

PROBLEMS THEV RDDRESS 

Inferring situation descriptions from sensor data 

Inferring likely consequences of given situations 

Inferring system malfunctions from observables 

Configuring objects under constraints 

Designing actions 

Comparing observations to plan vulnerabilities 

Prescribing remedies for malfunctions 

Executing a plan to adminiter a prescribed remedy 

Diagnose, debug, and repair student behavior 

Interpret, pred., repair, and monit. syst. behavior 



12 

Maynard (1934) proposed two general concepts of indus-

trial workplace design. The first concept was to reduce all 

motions used in the performance of the task to the lowest 

possible class. He defined five general classes ranging 

from the highest: finger, wrist, forearm, upper arm, and 

body motion to the lowest class which involved only finger 

motion. Maynard's second concept was to define the normal 

and maximum work areas in the horizontal and vertical 

planes. The normal working area in the horizontal plane was 

determined by arcs drawn with a sweeping motion of the arms. 

Only the forearms were extended, and the upper arms hung at 

the sides of the body. The maximum horizontal work area was 

determined in a similar fashion with the arms fully extend-

ed. The normal work area in the vertical plane included the 

area defined by arcs in the sagittal plane drawn by the 

forearms hinged at the elbows. Similarly, the maximum ver-

tical work area was determined by drawing arcs with the arms 

fully extended and hinged at the shoulders. The workplace 

layouts provided by Maynard were dimensionless and, there-

fore, of little use to a designer until Barnes (1940) 

applied dimensions to Maynard's layouts. 

Woodson and Conover (1964) recommended that anthropo-

m 
etric measurements of the largest worker should be used to 

determine workplace clearances, while those of the smallest 

worker should be used to determine limits of reach. They 

also cautioned the designer to make allowances for clothing, 



13 

which add to the clearance requirements and cause restric-

tion of movement. 

Konz (1967) conducted an experiment to investigate the 

effects of work surface heights on performance. Konz con-

cluded that the best working height for a standing operator 

is about 2.5 cm below the elbow. Ayoub (1973) recommended 

work surface heights for the standing and seated operator as 

a function of the work classification. He proposed three 

classifications of standing work and four classifications of 

seated work. Grandjean (1982) has made similar work surface 

height recommendations. 

V In order to design a successful industrial workplace, 

Ayoub (1973) encouraged designers to follow ten ergonomic 

guidelines: 

1. Reduce the static component of work. 

2. Do not overload the muscular system. 

3. Strive to achieve the best mechanical advantage. 

4. Eliminate extreme positions of the joints. 

5. Reduce unusual and stressful postures. 

6. Maintain a good seating arrangement. 

7. Permit change of posture on the job. 

8. Accommodate large operators and give them enough space. 

9. Train the operator to use the facilities. 

10. Match operator capacities with the job demands. 

In addition, Ayoub et al. (1982) examined the situa-

tions in which the operator should sit at the workplace, 



14 

stand at the workplace, or alternately sit and stand at the 

workplace. 

A. Seating is recommended at the workplace when: 

(1) All items needed in the short-term task cycle 
can be easily supplied and handled within the 
seated workspace. 

(2) The items being handled do not require the 
hands to work at an average level of more than 
15 cm above the work surface. 

(3) No large forces are required, such as handling 
weight greater than 4.5 kg. 

(4) Fine assembly or writing is done for a majori-
ty of the shift. 

B. Standing is recommended at the workplace when: 

(1) The workplace does not have knee clearance for 
a seated operator. 

(2) Objects weighting more than 4.5 kg are 
handled. 

(3) High, low, or extended reaches, such as those 
in front of the body, are required frequently. 

(4) Operations are physically separated and 
require frequent movement over a large area. 

(5) Downward forces must be exerted, as in packing 
and wrapping operations. 

C. Sit/Stand workplaces are considered in these 
instances: 

(1) Repetitive operations are done with frequent 
reaches more than 41 cm forward and/or more 
than 15 cm above the work surface. 

(2) Multiple tasks are performed, some best done 
sitting and others best done standing. 



CHAPTER II 

SYSTEM CONCEPTUALIZATION 

In order to conceptualize a complex, knowledge-based 

workplace design system one must establish many design goals 

and ideals. In the case of the subject workplace design 

system, goals and ideas were chosen to outline the system's 

potential functionality, usability, and accuracy. Develop-

ing the ideal workplace design system would involve 

hardware, expertise, time, and resources beyond the scope of 

a master's thesis project; however, the aim of this project 

is to conceptualize an ideal system and outline the proce-

dures required to develop the system. After performing the 

essential research and ground work required to begin 

development of the system, a subset of the conceptualized 

system was selected for prototyping. The working prototype 

can be tested and can serve as a basis for further research 

and development toward the completion of the ideal system. 

System Components 

The conceptualized workplace design system referred to 

above consists of three main components. These components 

are independent in a sense that they are not all necessary 

to build a workplace design system; yet the integration of 

all three components will provide the optimum environment 

for development of the ideal system. The three system com-

15 



16 

ponents are listed below. 

1. Intelligent Software 

2. Symbolic Processing Hardware 

3. Interactive Graphic Interface 

The intelligent software component of the system 

ensures that, during the design phase, the user adheres to 

accepted ergonomic concepts and principles. The software's 

functionality and interactiveness are such that even a 

novice unfamiliar with workplace design criteria can use the 

system and achieve optimum results. The inherent intelli-

gence of the software aids the user in solving problems 

requiring human judgment, experience, reasoning, and/or 

expertise. 

The intelligence of the software lies in the set of 

knowledge-based rules contained in the system. These rules 

allow the system to reason and to "think" like an expert in 

the field of workplace design. The system can interact with 

the user, make conclusions during the consultation session, 

and respond by providing an ergonomically sound workplace 

design. In addition, the system can justify the design for 

the user by recalling the internal knowledge that the con-

clusions were based on. 

The software acts as an aid to the workplace design 

engineer. By modeling the engineer's particular workplace 

characteristics and associated constraints, the software can 



17 

provide expert advice early in the design phase. The advice 

can range from suggestions to increase mechanical advantage 

to an actual layout drawing of the proposed workplace facil-

ity. Use of the system will ensure that ergonomic problems 

do not arise during the workplace implementation phase. 

This intelligent consulting service can not only save design 

time, but it can save money by eliminating design errors. 

The second component of the ideal workplace design sys-

tem is symbolic processing hardware. This type of computer 

hardware enables optimum, efficient functioning of intelli-

gent software coded in symbolic languages such as LISP or 

PROLOG. Conventional computers excel at problems that can 

be expressed in numerical terms and which lend themselves to 

repetitive, algorithmic solutions. However, traditional 

computing has not been effective in dealing with unstruc-

tured problems, interpreting information, using rules of 

thumb gained by experience, or dealing with uncertain or 

incomplete information. Symbolic processing is a technique 

that has been developed to address these problems. It 

refers to the utilization by computers of information and 

knowledge represented by symbols, analogous to the way 

humans reason with knowledge they possess. 

The special features that set symbolic processing com-

puters apart from conventional computers typically include a 

dedicated processor designed to optimize execution of sym-

bolic code, large virtual address space, large amounts of 



18 

physical memory, a high-resolution graphics display, and a 

high-performance mass storage device. Symbolic* processing 

hardware might be customized for the ideal workplace design 

system by reducing some of the knowledge sets to microchips, 

which might be inserted in the system's hardware. This 

would increase the speed and efficiency of the system. 

At present, no computer system exists that could be 

easily customized as a dedicated workplace design system. 

However, there are now computer systems available that util-

ize symbolic processing technology. Some of these systems 

include the Texas Instruments Explorer, the Xerox 1100, the 

Symbolics 3600, the LMI Lambda, and the Apollo system. 

An interactive graphic interface is the third component 

of the ideal workplace design system. Conceptually, this 

device enables a graphic interactivity between the system 

and the user. The primary graphic output of the system is a 

dimensioned drawing of the workplace layout. A side view, a 

top view, and a three-dimensional representation showing how 

the operator fits into the workplace facility is included. 

After the system has provided the initial design, the user 

can customize or modify the design, based on any physical 

constraints he may be confronting. If any significant 

changes are made to the design, the system responds by high-

lighting the problem areas such as reduced thigh clearance 

or reduced knee clearance. The severity of the problem is 

noted graphically and any consequential reduction in the 



19 

accommodated work force population will be indicated. 

The Context Tree 

The conceptualized ideal system can be pictured as a 

tree of interlocking knowledge sets or contexts. Texas 

Instruments (1985) defines a context as a structural unit 

used to separate the information in a knowledge base accord-

ing to main concepts or parts. If a knowledge base is 

thought of as a file cabinet full of information pertaining 

to one subject area or domain, the contexts are the drawers 

that are organized according to major divisions of informa-

tion. Contexts have parameters associated with them that 

store pieces of information used in the knowledge base. A 

special set of parameters, called goals, is attached to each 

context. After the values of all the goal parameters have 

been found, the context is exited. 

The context tree for the conceptualized workplace 

design system is presented in Figure 1. In a context tree a 

parent context is the context one level higher than the 

child context. The parent context of the subcontext 

entitled Physiological Design is the subcontext entitled 

Biomechanical Design. The root context. Dimensional Design, 

has no parent. The context called Other Ergonomic Consider-

ations has no child context but has the root context as its 

parent. A context's ancesto-rs include all the contexts that 

are in the direct line to the root context and including the 



20 

D I M E N S i O N R L 

D E S I G N 

B I O M E C H R N I C R L 

D E S I G N 

OTHER E R G O N O M I C 

C O N S I D E R R T I O N S 

P H V S I O L O G I C R L 

D E S I G N 

FIGURE 1 . The C o n t e x t T r e e 



21 

root context. A child context automatically inherits all 

the parameters of its ancestor contexts. When a context 

inherits the parameters of another context, it means that it 

can share the use of those parameters with the other con-

text . 

The root context. Dimensional Design, has many goal 

parameters associated with it. These include the type of 

workstation, the work surface height range, the chair height 

range, the footrest height range, the normal and maximum 

vertical reach, and the normal and maximum horizontal reach. 

In order to ascertain these values the system uses informa-

tion such as the defined work force population, 

anthropometric measurements from that population, the class-

ification of the work to be performed at the workplace, the 

physical constraints of the workplace, and the mathematical 

model developed by the author. 

The Biomechanical Design context more closely examines 

the dynamic components of the work. Factors such as the 

number of objects to be handled, object size, object weight, 

frequency of lift or movement, vertical or horizontal dis-

tance of movement, and distance between the object and the 

body are analyzed in this context. Goals for this context 

include the acceptable production rate at the workplace, the 

efficient arrangement of objects within the workplace, and 

any restrictions on the work force pertaining to operator 

fitness, muscular strength, or age. (An expert system that 



22 

analyses manual lifting tasks has been developed by 

Karwowski et al. [1985].) The purpose of this context is to 

provide optimum mechanical advantage for the operator and to 

ensure that he is in no immediate danger of over-exertion, 

injury, or disability. 

The Physiological Design context examines the energy 

requirements of the task to be performed at the workplace. 

Based upon a description of the motion components of the 

task and information gained from ancestor contexts, this 

context calculates the net energy requirements of the task 

in kilocalories per minute. Several models (Garg et al. 

1978, Asfour 1980) have been developed that enable this type 

of analysis. An evaluation of worker capacity is also 

included in this context. The user can rely on the internal 

physiological capacity data base (Astrand 1977) or he can 

enter data collected from his specific work force popula-

tion. The goal parameters for this context include the 

expected endurance time at this task, the required resting 

time, and the recommended number of rest periods for an 

eight hour shift. 

The final context. Other Ergonomic Considerations, is 

designed to provide informational guidelines that should be 

followed during the workplace design and implementation 

phases. This context provides guidelines concerning design 

of the interface between the operator and the workplace. 

Advice about the machine controls, displays, warning indica-



23 

tors, and other ergonomic and safety factors are presented. 

In addition, this context is knowledgable in national and 

international safety standards regarding the handling of 

toxic, hazardous, or radioactive materials; heat stress; 

cold stress; noise; lighting; and vibration. 

System Prototype 

After conceptualization of the ideal workplace design 

system, a subset of the system was selected for prototyping. 

Neither the equipment nor the expertise to customize symbol-

ic processing hardware or to design an interactive graphic 

interface device was immediately available. Therefore, it 

was decided that development of the system's intelligent 

software would be pursued. This pursuit was aided by the 

availability of a Texas Instruments Portable Professional 

Computer and the Personal Consultant Expert System Develop-

ment Tool. The system's root context. Dimensional Design 

(as shown in Figure 1), was the subset selected for proto-

typing. Subsequent chapters of the thesis describe the data 

that was used, outline the knowledge engineering activities 

that were performed, and detail the development of the pro-

totype knowledge base and the use of its inference engine. 



CHAPTER III 

KNOWLEDGE ENGINEERING 

Knowledge engineering is perhaps the most important 

phase of developing an expert system. Knowledge engineers 

acquire knowledge from a human expert and other sources and 

then embed that knowledge in an expert system. The process 

of knowledge engineering involves knowledge acquisition and 

knowledge representation. Knowledge is acquired through an 

extensive interview with a human expert to determine the 

thought processes he has when he solves a problem in the 

defined domain. Knowledge is represented by modeling those 

thought processes in the form of a knowledge base which in 

turn can be used by an inference engine to artificially 

reach the same solutions. This chapter describes the know-

ledge engineering procedures that were used to develop a 

workplace design expert system. 

Knowledge Acquisition 

The primary source of knowledge was Professor Ayoub, an 

experienced workplace design expert. In order to acquire 

his knowledge, it was necessary to discuss his methods of 

approaching an industrial workplace design project. Inter-

view questions were designed to establish the initial data 

that is required to begin a design project and what the out-

put of such a project should consist of. In addition, 

example design projects were presented to Professor Ayoub so 

24 



25 

that he could apply his expertise to these problems and 

reach satisfactory solutions. His knowledge and thought 

processes were examined in a step by step fashion as he made 

intermediate conclusions in solving the example projects. 

It is most helpful to have the human expert state his 

knowledge and thought processes in the form of IF/THEN 

statements. Consequently, the expert's knowledge can be 

acquired in a systematic, logical manner. If the expert's 

knowledge is acquired in this manner, it is relatively easy 

to represent his knowledge in a dynamic knowledge base that 

can be utilized by the inference engine. 

It was established that anthropometric measurements 

from the population to be accommodated at the workplace are 

the most important initial data required to begin a design. 

In addition, the classification of the work to be performed 

at the workplace is important, whether it is precision work, 

light work, heavy work, or VDT/keyboard operation. In order 

to establish what type of workstation to design (sit, stand, 

or sit/stand combination) it is necessary to find out if the 

operator is required to use foot controls, if he is required 

to cover a large work area, and to specify his reach 

requirements in terms of distance and frequency. Finally, 

it is important to decide what range of operators will be 

accommodated at the workplace. It is decided that a realis-

tic goal of a workplace design expert system is to design 

for a range that would accommodate the U. S. industrial pop-



26 

ulation from the 5th to the 95th percentiles, unless the 

design is restricted by physical constraints or specific 

anthropometric measurements entered by the user. 

Knowledge Representation 

Once the workplace design knowledge had been acquired 

it was necessary to represent the knowledge in a usable 

form. Thought processes were modeled by logical rules 

called modus ponens rules. The application of these rules 

is such that when A is known to be true and if a rule 

states, "If A, then B," it is valid to conclude that B is 

true. Stated differently, when we discover that the premis-

es of a rule are true, we are entitled to believe the 

conclusions. Modus ponens is the basic inferencing strategy 

used by most expert systems. 

In addition, most expert systems apply these rules to 

knowledge represented as pbject-attribute-value (0-A-V) 

triplets. In the case of the workplace design expert sys-

tem, the object is the root context, Dimensional Design. 

Attributes are equivalent to parameters. An example of a 

parameter might be Chair-Height. Values are the text or 

numerical denotations assigned to parameters. For example, 

the parameter, Chair-Height, might have the value 49.5 

centimeters. 0-A-V triplets are used to store informa-

tion in a knowledge base, while rules are used to infer 

conclusions from the knowledge base. 



27 

In order to represent the knowledge required to make a 

decision on what type of workstation to design, the set of 

rules presented in Table 3 were developed. The outcome of a 

rule depends on the value of the work classification para-

meter, the foot controls parameter, and the large work area 

parameter. In general, the aim of this set of rules is to 

conclude the workstation type as a sit/stand combination 

unless restricted by one or more of the premise parameters. 

The knowledge required to determine the physical dimen-

sions of the workplace is represented by another set of 

rules. In order to discuss these rules, the following para-

meters must be defined. 

Goal Parameters: 

WSU - the upper work surface height; 

WSL - the lower work surface height; 

CHU - the upper chair height; 

CHL - the lower chair height; 

FRU - the upper footrest height; 

FRL - the lower footrest height; 

NVR - the normal vertical functional reach; 

MVR - the maximum acceptable vertical reach; 

NHR - the normal horizontal functional reach; 

MHR - the maximum acceptable horizontal reach; 

Other Parameters: 

SH0ULDER5 - the smallest standing shoulder height; 



28 

TABLE 3. Rules for Determining the Workstation Type 

IF: 

Work-Class Foot-Controls 

Precision yes 

Precision 

Precision 

Precision 

P Light-Work 
1 

Light-Work 

Light-Work 

Light-Work 

Heavy Work 

Heavy Work 

Heavy Work 

Heavy Work 

VDT/Keyboard 

VDT/Keyboard 

VDT/Keyboard 

VDT/Keyboard 

yes 

no 

no 

yes 

yes 

no 

no 

yes 

yes 

no 

no 

yes 

yes 

no 

no 

Lar^e-Area 

yes 

no 

no 

yes 

yes 

no 

no 

yes 

yes 

no 

no 

yes 

yes 

no 

no 

yes 

THEN: 

Station-Tvoe 

Sit/Stand 

Sit 

Sit/Stand 

Sit/Stand 

Sit/Stand 

Sit 

Sit/Stand 

Sit/Stand 

Sit/Stand 

Sit/Stand 

Stand 

Stand 

Sit/Stand 

Sit 

Sit 

Sit/Stand 

/ 

/ 

/ 

/ 

^ 

/ 

/ 

/ 

/ 

/ 

-^ 



29 

SH0ULDER95 - the largest standing shoulder height; / 

SITTING-SH0ULDER5 - the smallest dimension from 

buttocks to shoulder; 

B0DY-DEPTH95 - the largest body depth dimension; 

THIGH-WIDTH5 - the smallest thigh clearance dimension; 

THIGH-WIDTH95 - the largest thigh clearance dimension; 

F0REARM5 - the smallest forearm length; /̂ 

ARM5 - the smallest arm length; ^ 

ELB0W5 - the smallest standing elbow height; J 

ELB0W95 - the largest standing elbow height; 

SITTING-ELB0W5 - the smallest dimension from buttocks 
to elbow; 

SITTING-ELB0W95 - the largest dimension from buttocks 

to elbow; 

P0PLITEAL5 - the smallest popliteal height; 

P0PLITEAL95 - the largest popliteal height; 

CI - correction factor applies to a Sit work surface; 
C2 - correction factor applies to Stand or Sit/Stand '^ 

work surfaces. 

The aim of the following set of rules is to accommodate 

the entire workforce population range at the workplace. It 

must be kept in mind that an operator with the smallest 

measurement for one parameter does not necessarily have the 

smallest measurement for the other parameters. For example, 

an operator might have the smallest popliteal height but the 

largest body-depth dimension. Therefore, these rules were 

developed with this knowledge in mind. The first rule 

applies to all workplaces. 



30 

If: STATION - TYPE = SIT OR STAND OR SIT/STAND 

Then: NHR = F0REARM5 - 1/2 (B0DY-DEPTH95) 

MHR = ARM5 - 1/2 (B0DY-DEPTH95) 

The next group of rules apply to fully adjustable work-

places in which the work surface height, chair height, and 

footrest height are all adjustable. 

If: STATION-TYPE = SIT 

Then: WSU = P0PLITEAL95 + SITTING-ELB0W95 + CI 

WSL = P0PLITEAL5 + SITTING-ELB0W5 + CI 

CHU = P0PLITEAL95 

CHL = P0PLITEAL5 

FRU = CHU - P0PLITEAL5 

FRL = 0 

NVR = CHL + SITTING-ELB0W5 + F0REARM5 

MVR = CHL + SITTING-SH0ULDER5 + ARM5 

I If: STATION-TYPE = STAND 

Then: WSU = ELB0W95 + C2 

WSL = ELB0W5 + C2 

CHU, CHL, FRU, FRL = NONE 

NVR = ELB0W5 + F0REARM5 

MVR = SH0ULDER5 + ARM5 

If: STATION-TYPE = SIT/STAND 

Then: WSU = ELB0W95 + C2 

WSL = ELB0W5 + C2 

CHU = ELB0W95 - SITTING-ELB0W5 



31 

CHL = ELB0W5 - SITTING-ELB0W95 

FRU = CHU - P0PLITEAL5 

FRL = 0 

NVR = ELB0W5 + F0REARM5 

MVR = SH0ULDER5 + ARM5 

The next group of rules apply to workplaces in which 

a workplace that includes a chair with no height adjustment 

feature. In order to accommodate the entire range of opera-

tors, the work surface height and the footrest height must 

be adjustable. Since the chair height is stationary, the 

CHU parameter will hold the value for the fixed chair height 

and the CHL parameter will be neglected. 

If: STATION-TYPE = SIT 

Then: CHU = P0PLITEAL95 

CHL = NONE 

WSU = CHU + SITTING-ELB0W95 + CI 

WSL = CHU + SITTING-ELB0W5 + CI 

FRU = CHU - P0PLITEAL5 

FRL = 0 

NVR = CHU + SITTING-ELB0W5 + F0REARM5 

MVR = CHU + SITTING-SH0ULDER5 + ARM5 

If: STATION-TYPE = STAND 

Then: CHU, CHL = NONE 

WSU = ELB0W95 + C2 



32 

WSL = ELB0W5 + C2 

FRU, FRL = NONE 

NVR = ELB0W5 + F0REARM5 

MVR = SHOULDERS + ARM5 

If: STATION-TYPE = SIT/STAND 

Then: CHU = ELB0W95 - SITTING-ELBOWS 

CHL = NONE 

WSU = CHU + SITTING-ELB0W95 + C2 

WSL = CHU + SITTING-ELB0W5 + C2 

FRU = CHU - P0PLITEAL5 

FRL = 0 

NVR = ELB0W5 + FOREARMS 

MVR = SHOULDERS + ARMS 

The next group of rules apply to workplaces in which 

the work surface height is fixed. These rules enable the 

design of a workplace that includes a work surface or table 

with no height adjustment feature. In order to accommodate 

the entire range of operators, the chair height and the 

footrest height must be adjustable. Since the work surface 

height is stationary, the WSU parameter will hold the value 

for the fixed work surface height, and the WSL parameter 

will be neglected. For the cases in which the workstation 

type is stand or sit/stand, the footrest dimensions refer to 

an adjustable platform on which the operator can adjust him-

self to the fixed work surface. Determining workplace 



33 

dimensions for this group of rules is more difficult because 

the relative position of the elbow, buttocks, or foot must 

be solved for in relation to the fixed work surface. 

If: STATION-TYPE = SIT 

Then: WSU = P0PLITEAL9S + SITTING-ELB0W95 + CI 

WSL = NONE 

Equating at the elbow, 

CHU + SITTING-ELBOWS = WSU - CI 

and by substitution, 

CHU = P0PLITEAL9S + SITTING-ELB0W9S -
SITTING-ELBOWS 

CHL = P0PLITEAL9S 

FRU = CHU - POPLITEALS 

FRL = 0 

Equating at the elbow, 

NVR - FOREARMS = WSU - CI 

and by substitution, 

NVR = P0PLITEAL95 + SITTING-ELB0W95 + 
FOREARMS 

Equating at the buttocks, 

MVR - SITTING-SHOULDERS - ARMS = CHL 

and by substitution, 

MVR = P0PLITEAL9S + SITTING-SHOULDERS + 
ARMS 

If: STATION-TYPE = STAND 

Then: WSU = ELB0W9S + C2 

WSL = NONE 



34 

CHU, CHL = NONE 

Equating at the foot, 

FRU = WSU - C2 - ELBOWS 

and by substitution, 

FRU = ELB0W9S - ELBOWS 

FRL = 0 

Equating at the elbow, 

NVR - FOREARMS = WSU - C2 

and by substitution, 

NVR = ELB0W95 + FOREARMS 

Equating at the foot, 

MVR - SHOULDERS - ARMS = WSU - C2 - ELB0W9S 

and by substitution, 

MVR = SHOULDERS + ARMS 

STATION-TYPE = SIT/STAND 

Then: WSU = ELB0W9S + 0 2 " ^ 

WSL = NONE 

Equating at the elbow, 

CHU + SITTING-ELBOWS = WSU - C2 

and by substitution, 

.CHU = ELB0W9S - SITTING-ELBOWS 

Similarly, 

CHL = ELB0W95 - SITTING-ELB0W9S 

FRU = CHU - POPLITEALS 

FRL = 0 



35 

Equating at the elbow, 

NVR - FOREARMS = WSU - C2 

and by substitution, 

NVR = ELB0W9S + FOREARMS 

Equating at the buttocks, 

MVR - SITTING-SHOULDERS - ARMS = CHL 

and by substitution, 

MVR = ELB0W9S - SITTING-ELB0W95 + 
SITTING-SHOULDERS + ARMS 

The next group of rules apply to workplaces in which 

all dimensions are fixed. These rules enable the design of 

a workplace that includes a work surface, chair, and foot-

rest with no height adjustment features. Obviously, this 

type of design cannot accommodate the entire range of opera-

tors. Specifically, a percentage of the workforce 

population will be excluded from the workplace due to thigh 

clearance restrictions. In addition, closer examination of 

the previous rules indicate that possible restrictions 

could result depending on the thickness of the work surface 

A set of rules that calculates the severity of the thigh 

clearance restriction will be presented after the following 

group of rules. Since the height dimensions of the follow-

ing three rules are stationary, the WSU, CHU, and FRU 

parameters will hold the values for those fixed dimensions, 

and the WSL, CHL, and FRL parameters will be neglected. 



36 

If: STATION-TYPE = SIT 

Then: WSL, CHL, FRL = NONE 

WSU = P0PLITEAL95 + SITTING-ELB0W9S + CI 

CHU = WSU - CI - 1/2 (SITTING-ELBOWS + 
SITTING-ELB0W9S) 

FRU = CHU - 1/2 (POPLITEALS + P0PLITEAL95) 

NVR = CHU + SITTING-ELBOWS + FOREARMS 

MVR = CHU + SITTING-SHOULDERS + ARMS 

If: STATION-TYPE = STAND 

Then: WSL, CHL, FRL = NONE 

WSU = ELB0W9S + C2 

CHU = NONE 

FRU = WSU - C2 - 1/2 (ELBOWS + ELB0W9S) 

NVR = FRU + ELBOWS + FOREARMS 

MVR = FRU + SHOULDERS + ARMS 

If: STATION-TYPE = SIT/STAND 

Then: WSL, CHL, FRL = NONE 

WSU = ELB0W95 + C2 

CHU = WSU - C2 - 1/2 (SITTING-ELBOWS + 
SITTING-ELB0W9S) 

FRU = CHU - 1/2 (P0PLITEAL5 + P0PLITEAL95) 

NVR = CHU + SITTING-ELBOWS + FOREARMS 

MVR = CHU + SITTING-SHOULDERS + ARMS 

The correction factors, CI and C2, are used to modify 

the work surface height based on the classification of the 



37 

work to be performed at the workplace. For example, preci-

sion work requires a work surface above the elbows and 

closer to the eyes, while heavy work requires a work surface 

below the elbows to allow exertion of greater forces. The 

rules that determine the CI and C2 values are listed below. 

If: WORK-CLASS = PRECISION-WORK 

Then: CI = +12 cm. 

C2 = +8 cm. 

If: WORK-CLASS = LIGHT-WORK 

Then: CI = +5 cm, 

C2 = -5 cm. 

If: WORK-CLASS = HEAVY-WORK 

Then: CI = 0 cm. 

C2 = -13 cm, 

If: WORK-CLASS = VDT/KEYBOARD-OPERATION 

Then: CI = 0 cm. 

C2 = -5 cm. 

As discussed previously, the completely stationary 

design and the other adjustable designs present possible 

thigh clearance restrictions. The next set of rules deter-

m ines what portion of the work force population will be 

excluded from the workplace due to a thigh clearance problem 

by calculating the approximate reduction percentage. In 

order to make this estimation, it is assumed that the 



38 

distribution of thigh clearance measurements over the popu-

lation range is normally distributed. Before these rules 

can be discussed, the following parameters must be defined, 

TABLE-THICKNESS - vertical thickness of the work surface; 

THIGH-TEST - determines the existence of a thigh clearance 

problem; 

THIGH-CUT - severity of the restriction in centimeters; 

THIGH-POPS - the smallest thigh width/popliteal 

combination; 

THIGH-P0P9S - the largest thigh width/popliteal combination; 

MU-HAT - the estimated distribution mean; 

SIGMA-HAT - the estimated distribution standard deviation; 

THIGH-CRITICAL - the calculated distribution Z value. 
If: DESIGN-TYPE = STATIONARY, and 

THIGH-WIDTH9S > (WSU - TABLE-THICKNESS) - CHU 

Then: THIGH-TEST = YES 

If: DESIGN-TYPE = FIXED-WORK-SURFACE, and 

STATION-TYPE = SIT/STAND, and 

THIGH-WIDTH9S > (WSU - TABLE-THICKNESS) - CHU 

Then: THIGH-TEST = YES 

If: THIGH-TEST = YES 

Then: THIGH-CUT = THIGH-WIDTH9S - (WSU -

TABLE-THICKNESS - CHU) 

THIGH-POPS = THIGH-WIDTHS + POPLITEALS 

THIGH-P0P95 = THIGH-WIDTH95 + P0PLITEAL9S 
MU-HAT = 1/2 (THIGH-POPS + THIGH-P0P9S) 



39 

SIGMA-HAT = (THIGH-P0P95 - MU-HAT) / 1.645 

THIGH-CRITICAL = THIGH-P0P9S - THIGH-CUT 

If: (THIGH-CRITICAL - MU-HAT) / SIGMA-HAT <= 1.645 

(THIGH-CRITICAL - MU-HAT) / SIGMA-HAT > 1.28 

Then: POPULATION-REDUCTION = < 5% 

If: (THIGH-CRITICAL - MU-HAT) / SIGMA-HAT <= 1,28 

(THIGH-CRITICAL - MU-HAT) / SIGMA-HAT > 1.035 

Then: POPULATION-REDUCTION = 5-10% 

If: (THIGH-CRITICAL - MU-HAT) / SIGMA-HAT <= 1.035 

(THIGH-CRITICAL - MU-HAT) / SIGMA-HAT > .0675 

Then: POPULATION-REDUCTION = 10-20% 

If: (THIGH-CRITICAL - MU-HAT) / SIGMA-HAT <= 0.675 

(THIGH-CRITICAL - MU-HAT) / SIGMA-HAT > 0.385 

Then: POPULATION-REDUCTION = 20-30% 

If: (THIGH-CRITICAL - MU-HAT) / SIGMA-HAT <= 0.385 

(THIGH-CRITICAL - MU-HAT) / SIGMA-HAT > 0.125 

Then: POPULATION-REDUCTION = 30-40% 

If: (THIGH-CRITICAL - MU-HAT) / SIGMA-HAT <= 0.125 

(THIGH-CRITICAL - MU-HAT) / SIGMA-HAT > 0 

Then: POPULATION-REDUCTION = 40-50% 

If: (THIGH-CRITICAL - MU-HAT) / SIGMA-HAT <= 0 

Then: POPULATION-REDUCTION = >50% 



CHAPTER IV 

DEVELOPMENT OF THE SYSTEM 

Once the knowledge for the Dimensional Design context 

was acquired and represented in rule form, it was possible 

to begin development of the system prototype. A personal 

computer-based expert system building tool was used to 

assist in the development of the system. The tool chosen 

was the Personal Consultant Expert System Development Tool. 

Personal Consultant is implemented in the IQLISP computer 

language. In other words, the inference engine and the 

knowledge engineer/user interfaces are coded in the underly-

ing IQLISP language. The tool does not offer a compiler to 

make completed programs faster to operate. However, the 

existing configuration is quite adequate for the development 

of an expert system. Personal Consultant is composed of a 

development engine and an inference engine. The development 

engine enables the knowledge engineer to develop and main-

tain the knowledge base. The inference engine enables the 

knowledge engineer to test the knowledge base, and it 

enables the user to execute a consultation. The architec-

ture of a completed expert system is presented in Figure 2. 

The Development Engine 

The development engine is an interactive, window-

oriented interface to the knowledge engineer that helps him 

to build his knowledge base into an expert system. In 

40 



41 

K N O U I L E D G E BRSE 

RULES FRCTS 

I N F E R E N C E E N G I N E 

I N F E R E N C E CONTROL 

— z — 

D E U E L O P M E N T 

E N G I N E 

L D O R K I N G 

M E M O R V 

1 
1 
1 
1 

1 
1 
1 
1 
1 
1 

1 
1 

EHPERT OR 

K N O U I L E D G E 

E N G I N E E R 

USER 

FIGURE 2 . T h e A r c h i t e c t u r e of a n E x p e r t S y s t e m ( f r o m Harmon 
a n d K i n g 1 9 8 5 ) 



42 

addition, the development engine's interactive structure 

editor allows on-line modification of an existing knowledge 

base, A knowledge base is entered by responding to prompts. 

The first prompt asks for the name of the knowledge base 

heading or domain. Then values for root context properties 

are prompted for, including a list of initial data and goal 

parameters. Once these parameter names have been entered, 

the development engine asks for information about the para-

meters. Next, the context tree must be described by 

entering the descendants of the root context. At this 

point, rules can be inserted into the knowledge base. Rules 

can be entered directly in LISP or in an Abbreviated Rule 

Language (ARL). ARL is a BASIC-like rule specification lan-

guage that simplifies the rule specification procedure for 

those unfamiliar with the LISP computer language. During 

the definition of the rules, if a parameter name is entered 

that has not been specified, the development engine prompts 

for information about the parameter. In this manner con-

text, parameter, and rule properties are specified by the 

knowledge engineer until the knowledge base is completed. 

Definitions of knowledge base properties are explained in 

Appendix A. 

The main activity menu of the development engine pre-

sents a list of 13 options designed to help modify or debug 

a knowledge base. The first option. Go, begins a consulta-

tion with the specified knowledge base, and the second 



43 

option, Quit, ends a Personal Consultant session. The third 

option, Lisp, allows the knowledge engineer to activate the 

IQLISP interpreter if he wants to write special operations 

to augment the knowledge base. The Parameters, Rules, Con-

texts, and Variables options allow properties associated 

with these entities to be entered or modified. The Func-

tions option enables the knowledge engineer to define 

special functions written in LISP that are not immediately 

available in the development engine. The List option pro-

vides a complete listing of the knowledge base. The Save 

option writes the knowledge base to a computer file. The 

Trace option is a debugging utility that produces a listing 

of the rules applied and the parameter values used in a 

given cojisultation. The final two options. Record and Play-

back, are used to record a set of prompts with their 

respective responses from a given consultation and then use 

those responses in an automated session to detect changes in 

the logical flow of the knowledge base. 

Description of the Data 

A conventional computer program that accesses a data 

base usually stores the data in an external data file or in 

internal arrays. However, data used by an expert system can 

be inserted directly into the knowledge base with the aid of 

the development engine. Numerical data can be included as 

parameter values in rules that are needed to access the 



44 

data. 

The data base used by the workplace design expert sys-

tem is a collection of anthropometric measurements taken 

from the U. S. military population (Hertzberg 1972). The 

data were derived on the basis of measurements from nude 

subjects. Therefore, the data were duly adjusted for cloth-

ing, shoe, and other allowances which were added or 

subtracted to the various measurements. For the standing 

shoulder height, 2.5 cm were added for men's and women's 

shoes and 1.4 cm were added for clothing. The sitting 

shoulder height required an additional 0,6 cm for clothing 

under the buttocks. 1.0 cm was added to the body depth 

measurement and 2.0 cm were added to the thigh width 

measurement for clothing allowances. For the forearm 

length, 7.6 cm were subtracted for thumb and forefinger 

manipulation. For the arm length, 5.1 cm were subtracted to 

account for measurement from the back of the shoulder and 

7,6 cm were subtracted for thumb and forefinger manipula-

tion, 2,5 cm were added for men's and women's shoes to the 

standing elbow height. No allowances were given for the 

sitting elbow height and 2.5 cm were added to the popliteal 

height for men's and women's shoes (Hertzberg 1972). The 

corrected anthropometric data base is presented in Table 4. 

The workplace design expert system user can select the 

population that he wishes to accommodate at his workplace. 

The valid populations are male, female, and mixed. If the 



45 

TABLE 4. Corrected Anthropometric Measurements (from 
Hertzberg 1972, Das and Grady 1983) 

PARAMETER 

Shoulder 5 

Shoulder 95 

S i t t i n g - S h o u l d e r 5 

B o d y - D e p t h 

T h i g h - l U i d t h 

Thigh-LDidth 

F o r e a r m 5 

Rrm 5 

Elboui 5 

Elbouj 95 

S i 11 i n g - E1 b 0 uj 

S i 11 i n g - E1 b 0 uj 

P o p l i t e a l 5 

P o p l i t e a l 95 

95 

5 

95 

5 

95 

Male 

1 3 8 . 0 

1 5 6 . 8 

5 4 . 7 

3 4 . 0 

1 4 . 2 

1 8.5 

3 7 . 1 

6 8 . 3 

1 0 5 . 6 

1 2 0 . 4 

1 8.8 

2 7 . 4 

4 2 . 4 

4 8 . 7 

VALUE (cm) 

Female 

1 2 7 . 9 

1 4 6 . 6 

5 2 . 2 

2 7 . 6 

1 2 . 4 

1 6.5 

3 2 . 5 

6 0 . 2 

9 9 . 0 

1 1 1 . 2 

1 8.8 

2 7 . 4 

3 7 . 3 

4 3 . 6 



46 

male population is chosen, 95th percentile male anthropo-

metric measurements are used to determine clearance 

dimensions while Sth percentile male measurements are used 

to determine reach dimensions. The method is similar for 

the female population, /if a mixed population is chosen, 

9Sth percentile male measurements are used for clearance 

determinations, and Sth percentile female measurements are 

used to calculate the reach requirements. 

Construction of the Knowledge Base 

The workplace design expert system knowledge base was 

constructed using the development engine as described at the 

beginning of this chapter. The knowledge base heading pro-

perty was specified to give the knowledge base a name. 

Henceforth, the knowledge base will be referred to by its 

name, "The Ergonomist." 

All of the parameters used by "The Ergonomist" were 

defined in terms of their associated properties. See Appen-

dix A for an explanation of knowledge base properties. A 

total of 41 parameters were defined for use by "The Ergono-

mist," All parameters belong to the parameter group called 

Dimensional-Design-Farms. 

The workplace design knowledge represented in rule form 

and presented in Chapter III was entered directly into the 

knowledge base with the aid of the development engine. 

Since the knowledge was already organized as IF/THEN state-



47 

ments, the rules were entered with very little modification 

A total of 57 rules were embedded in "The Ergonomist" and 

were grouped as follows: 6 Anthro-Rules, 16 Station-Type-

Rules, 16 Dimensioning-Rules, 12 Thigh-Clearance-Rules, and 

7 Data-Checking-Rules. 

The Inference Engine 

"The Ergonomist" knowledge base, when used in conjunc-

tion with the inference engine, comprises the functioning 

workplace design expert system prototype. The inference 

engine controls the expert system's reasoning process as it 

uses the facts and rules stored in the knowledge base, as 

well as the information it acquired from the user. The 

inference engine performs two major tasks (Harmon and King 

1985). First, it examines existing facts and rules, and 

adds new facts when possible. Second, it decides the order 

in which inferences are made. In doing so, the inference 

engine conducts the consultation with the user. 

To reach a conclusion about the goal parameters, the 

inference engine employs an inferencing strategy. Figure 3 

presents the inferencing strategies that are utilized by 

various inference engines. Backward and forward chaining 

strategies can be utilized, as well as depth-first and 

breadth-first search strategies. Backward chaining infer-

ence engines, or goal-directed systems, work backward 

through the knowledge base in an effort to choose an answer 



48 

p 

2 

& 
« 

(0 

OQ 

j 2 i n 
00 

0) I—I 
W 

3 O O 
C 

CO - H 
d) ^ 

• H 
0 0 T 3 
(U C 

4-> CO 
CO 
Ji C 
4-1 O 
CO e 

u 
x : CO 
a K 
u 
CO E 

cu o 
C O S-i 

M - l 
M - l > ^ 

O 
CO 

CO (1) 
CU C 

• H - H 
V̂  0 0 
O C 
O O W 
CU 
4-) (U 

a 
c 
( U 
M 
<U 

CO 

o 

CO C 

CO 

w 

o 
I—I 



49 

For example, if a goal parameter is D, and there is a rule 

that states "if C, then D" the inference engine will attempt 

to ascertain the value for C. The inference engine may dis-

cover another rule that states "if A and B, then C." Then 

the values for A and B must be found. This is also an 

example of a depth-first search strategy. A breadth-first 

search sweeps across all of the goal parameters before digg-

ing for greater detail. In cases in which the number of 

possible outcomes is large, a forward chaining strategy is 

often used. In a forward chaining system, or data-driven 

system, premises of rules are examined to see whether or not 

they are true, given the information available at the time. 

If so, then the conclusions are added to the list of facts 

known to be true and with the new information, the system 

examines the rules again. 

The inference engine supplied with the Personal 

Consultant employs a depth-first, backward chaining 

strategy. The inference engine controls the flow of a 

consultation with "The Ergonomist." Inference control is 

actually spread subtly throughout "The Ergonomist." since 

the order of clauses within a rule determines which clause 

is examined in what order. For example, there is a rule in 

the knowledge base that states, "if WORK-CLASS = PRECISION 

and FOOT-CONTROLS = YES and LARGE-AREA = NO, then STATION-

TYPE = SIT." As a result, the inference engine tries to 

find answers for those three premise clauses in the order in 



so 

which they appear. Therefore, a degree of inference control 

is contained in the rule structure. 

During a consultation, the user has the ability to 

query the system, A help key is provided that can help 

clarify the meaning of a question presented to the user. If 

the user does not understand the meaning of a question, he 

can use the help key for further clarification. The 

resulting help message is the value of a knowledge base 

parameter property called reprompt. The why key is used to 

clarify the logic being used by the expert system. For 

example, if the user is presented with a question, he has 

the ability to ask the system why the answer to that 

question is needed. The system responds by listing the rule 

being examined for which the answer is required in order to 

apply the rule. Finally, the how key enables the user to 

view the logic that was used to determine the value of 

parameters already found by the expert system. Use of this 

key lists a chain of rules that determined the value for a 

selected parameter. 

Sample consultations with "The Ergonomist" are 

presented in Appendix C. The sample consultations simulated 

actual workplace design projects that could be encountered 

in an industrial setting and solved by "The Ergonomist" 

expert system. 



CHAPTER V 

CONCLUSION 

The development of a workplace design expert system 

that acts as an intelligent aid to workplace designers has 

been described. The expert system will potentially benefit 

facility engineers, human factors engineers, safety engin-

eers, and industrial engineers. In addition, the system 

will be useful as an educational tool for students in those 

areas of specialization. 

The expert system will be most beneficial early in the 

design phase of a new workplace. Potential ergonomic prob-

lems with the design can be avoided at great savings in time 

and money. The system ensures that the workplace is 

designed to fit the operator; it does not force the operator 

to fit the workplace. Use of the expert system to assist in 

the design of workplaces should reduce incidents of over-

exertion and injury at the workplace. This result will in 

turn reduce compensation expenditures and avoid man hour 

losses . 

In addition to designing new workplaces, the expert 

system has the potential to evaluate existing workplaces. 

After the user has responded to questions asked about the 

existing workplace, the system will provide a description of 

what the workplace characteristics should be. Comparisons 

can be made to the existing workplace and necessary modifi-

cations can be identified. The system can be used 

51 



52 

iteratively to evaluate all existing workplaces in an indus-

trial plant. 

The effort that went into the completion of this thesis 

project serves as a foundation for further research and 

development, "The Ergonomist" is a prototype of the ideal 

workplace design system conceptualized in Chapter II. As 

better expert system building tools and symbolic processing 

hardware become available, the development of a workstation 

dedicated to the design of industrial workplaces will become 

a reality. 

The expert system development tool used to create "The 

Ergonomist" has many useful features; however, there are 

areas that need improvement. The tool does not permit the 

use of complex rules premises. For example, a rule premise 

such as "if A and (B or C or D)" is not valid. Premises 

must contain all "ands" or all "ors." In addition, it would 

be desirable to be able to use other logical operators such 

as "not," "nand," and "nor" which are not recognized by the 

tool. The tool does not provide a means for using a graphic 

user interface. It would be highly desirable to provide a 

graphical output of the finished workplace design. In addi-

tion, the meaning of certain questions asked by "The 

Ergonomist" could be clarified with graphics. For example, 

if the user is asked to enter the popliteal height measure-

ment, a graphic screen could indicate how this measurement 

should be taken on the image of an anatomical manikin. 



S3 

The Ergonomist" expert system has yet to undergo 

extensive testing in industry. Only the direct application 

of this system over an extended period of time will prove 

whether the system is beneficial in avoiding ergonomic prob-

lems and reducing injury at the workplace. 



LIST OF REFERENCES 

Aikens, J. S,, "Prototypes and production rules: A knowledge 
representation for computer consultations." 
Unpublished Ph, D. dissertation, Stanford University, 
1980, 

Asfour, S, S., "Energy cost prediction models for manual 
lifting and lowering tasks." Unpublished Ph, D. 
dissertation, Texas Tech University, 1980, 

Astrand, P. 0., Textbook of Work Physiology. McGraw-Hill, 
1977. 

Ayoub, M. M., J, Fernandez, J, Smith, "Design of workplace." 
Unpublished report, Texas Tech University, 1982. 

Ayoub, M. M., "Workplace design and posture." Human 
Factors, 15, No. 3, 265-268, 1973. 

Barnes, R. M., Motion and Time Study, 2nd Edition, Wiley, 
1940. 

Buchanan, B. G., E. Shortliffe, Rule-Based Expert Systems: 
The MYCIN Experiments of the Stanford Heuristic 
Programming Project, Addison-Wesley, 1984. 

Das, B., R. Grady, "Industrial workplace layout and 
engineering anthropology." In T. Kvalseth (ed.) 
Ergonomics of Workstation Design, Butterworth, 1983. 

Duda, R. 0., R. Reboh, "AI and decision making: The 
PROSPECTOR experience." In W. Reitman (ed.) Artificial 
Intelligence Applications for Business, Ablex, 1984. 

Erman, L. D., F. Hayes-Roth, V. Lesser, D. Reddy, "The 
HEARSAY II speech-understanding system: Integrating 
knowledge to resolve uncertainty." Computing Surveys, 
12 No. 2, 213-253, 1980. 

Feigenbaum, E. A., "The art of artificial intelligence: 
Themes and case studies of knowledge engineering." 
IJCAI 5, 1014-1029, 1977. 

Garg, A., D. Chaffin, G. Herrin, "Prediction of metabolic 
rates for manual materials handling jobs." Am. Ind. 
Hyg. Assoc. J., 13, 661-674, 1978. 

Grandjean, E., Fitting the Task to the Man: An Ergonomic 
Approach, Taylor and Francis, 1982. 

54 



55 

Harmon, P., King, D., Expert Systems: Artificial 
Intelligence in Business. Wiley, 1985. 

Hayes-Roth, F., D. Waterman, D. Lenat (eds.). Building 
Expert Systems, Addison-Wesley, 1983. 

Hertzberg, H. T. E., "Engineering anthropology." In Van 
Cott H., R, Kincade (eds,) Human Engineering Guide to 
Equipment Design. McGraw-Hill, 1972. 

Karwowski, W., N. Mulholland, T. Ward, V. Jagannathan, R. 
Kirchner, "LIFTAN: A knowledge-based expert system for 
analysis of manual lifting tasks." Unpublished report. 
University of Louisville, 1985. 

,X6nz, S., "Design of work stations." J. Industrial Eng. 18, 
No. 7, 413-423, 1967. 

Konz, S., Work Design. Grid, 1979. 

Lindsay, R. K., B. Buchanan, E. Feigenbaum, J. Lederberg, 
Chemistry: The DENDRAL Project. McGraw-Hill, 1980. 

Martin, W. A., R. Fateman, "The MACSYMA system." Proceedings 
of the 2nd Symposium on Symbolic and Algebraic 
Manipulation. 59-75, 1971. 

Maynard, H. B., "Workplace layouts that save time, effort 
and money." Iron Age 134, No. 23, 28-30, 1934. 

Smith, J. L., J. Ramsey, "Designing physically demanding 
tasks to minimize levels of worker stress." Industrial 
Engineering, May, 1982. 

Slagle, J. R., "A heuristic program that solves symbolic 
integration problems in freshman calculus." Symbolic 
Automatic Integrator (SAINT), Unpublished Ph. D. 
dissertation, MIT, 1961. 

Texas Instruments, Inc., "Personal Consultant: Expert system 
development tools." User's Guide, T. I., 1985. 

Tichauer, E. R., Occupational Biomechanics: The Anatomical 
Basis of Workplace Design, New York University Medical 
Center, No. 51, 1975. 

Utfoodson, W. E., D. W. Conover, Human Engineering Guide for 
Equipment Designers. 2nd Edition, University of 
California Press, 1964. 



APPENDIX A 

DEFINITION OF KNOWLEDGE BASE PROPERTIES 

This appendix contains explanations of the concepts and 

terminology used by the Personal Consultant in describing 

properties of a knowledge base. This information is 

provided so that structure and context of a knowledge base 

can be better understood. Refer to Appendix B, Listing of 

the Knowledge Base, for a detailed description of "The 

Ergonomist" knowledge base properties. A knowledge base 

consists of three main components: contexts, parameters, and 

rules. The properties pertaining to these components are 

explained below as defined by Texas Instruments (1985). 

Contexts 

* PARMGROUP - The name of the parameter group associated 
with a context. The parameter group contains the names 
of all parameters in the group. There is only one 
parameter group associated with a context. 

* RULETYPES - The names of all rule groups associated with a 
context. Like the parameter group, the rule group 
contains the names of all rules in the group. There 
can be several rule groups associated with a context. 

* GOALS - A list of the parameters in a context's parameter 
group whose values, when determined, describe the 
solution to the context problem. 

* DISPLAYRESULTS - A value that specifies if the value of 
the parameter(s) listed in the GOALS property is to be 
displayed when the context is exited. 

* INITIALDATA - A list of the parameters in the context's 
parameter group whose values are always prompted for 
when the context is instantiated during the 

56 



57 

consultation. 

* PROMPTEVER, PROMPTIST, PR0MPT2ND, TRANS - Text that is 
used to describe the type of information needed in the 
consultation. These properties are used by the infer-
ence engine to introduce or announce the instance of a 
context or describe the purpose of the context. 

* SYN, UNIQUE - Values that are used to translate the con-
text instance name into a meaningful form for the 
client. 

PRINTID - Property whose value is used to name a context 
instance. 

* ASSOCWITH, OFFSRING - Values used to store information 
about the context tree. 

Parameters 

* PROMPT - This property is usually the question that asks 
the client for the value of the parameter. If this 
property contains the question, the question is 
displayed to the client. 

* REPROMPT - This property is an explanation or help message 
for the prompt that is displayed to the client during a 
consultation. 

* TRANS - This property is an English phrase that describes 
the purpose or use of the parameter. This phrase is 
used to generate questions, explain the reasoning or 
inferencing of the inference engine, or describe the 
results of a consultation. Correct form in the TRANS 
property helps make a knowledge base easier to 
understand and to correct if problems exist. 

* ASKFIRST - This property tells the inference engine to ask 
the client for the value before trying to infer it 
using rules. 

* EXPECT - This property describes the types of input values 
expected from the client. This property can be any of 
the following: 

- A list of all appropriate values - the response can 
be any of the values in the list. 

- NIL - Expect a yes or no response from the client 



58 

- ANY - Any type of response is legal. 

- NUMB - The input value must be a number. 

- POSNUMB - The input value must be a number greater 
than or equal to zero. 

- FIXP - The input value must be an integer, 

LEGALVALS - This property describes all possible values of 
a parameter. When a parameter does not have a 
LEGALVALS property, the inference engine assumes that 
the EXPECT property specifies all the legal values. 
This property is required for multivalued parameters. 
This property can be a list of all legal values, or one 
of these special values: 

- ANY - Any value is a legal parameter value. 

- TEXT - The parameter's value can be any text phrase. 
Parameters with this LEGALVALS property value can be 
used to store text which can explain the conclusion 
or recommendations of a consultation. 

* MULTIVALUED - This property is used to define the type of 
a parameter. If this property does not exist, the 
parameter is single-valued. 

* USED-BY, UPDATED-BY, CONTAINED-IN, ANTECEDENT-IN, 
UPDATED-IN - These system properties are used to identify 

rules that use or modify the parameter's values. These 
properties are automatically maintained by the 
development engine. These properties are useful during 
the debugging process for the knowledge base. 

Rules 

* PREMISE - This property defines all the conditions to be 
met before the rule action will be taken. 

* ACTION - This property defines the actions taken if the 
premise is true. 

^ SUBJECT - This property is the rule group that the rule 
belongs to. This determines the context(s) in which 
the rule must be tried. 

•5̂  ANTECEDENT - This property is optional and indicates that 
the rule is a forward-chaining rule. 



APPENDIX B 

LISTING OF THE KNOWLEDGE BASE 

Rule Group ANTHRO-RULES 

RULEOOl [ANTHRO-RULES] 

If the workf 
Then 1) 

2) 

3) 

^ ) 

5) 

6) 

7) 

8) 

9) 

10) 

11) 

12) 

it i 
Stan 
it i 
Stan 
it i 
Stan 
It 1 
Stan 
it i 
popl 
it i 
popl 
It 1 
sitt 
It 1 
sitt 
it i 
sitt 
it 
bod 
it 
per 
it 
per 

orce 
s de 
ding 
s de 
ding 
s de 
ding 
s de 
ding 
s de 
itea 
s de 
itea 
s de 
ing 
s de 
ing 
s de 
ing 
is d 
y de 
is d 
cent 
is d 
cent 

populat 
finite ( 
elbow 

finite 
elbow 

finite 
shoulde 

finite ( 
shoulde 

finite ( 
1 height 
finite ( 
1 height 
finite ( 
elbow he 
finite ( 
elbow he 
finite ( 
shoulder 
efinite 
pth is 3 
efinite 
lie fore 
efinite 
ile arm 

ion IS 
100%) 
eight 
100%) 
eight 
100%) 
r heig 
100%) 
r heig 
100%) 
is 48 
100%) 
is 42 
100%) 
ight i 
100%) 
ight i 
100%) 
heigh 
(100%) 
4.0, a 
(100%) 
arm is 
(100%) 
is 68. 

MALE, 
that t 
is 120 
that t 
is 105 
that t 
ht is 
that t 
ht is 
that t 
. 7 , an 
that t 
. 4 , an 
that t 
s 27.4 
that t 
s 18.8 
that t 
t is 5 
that 

nd 
that 
37.1, 
that 

3. 

he 9 
.4, 
he 5 
.6, 
he 9 
156. 
he 5 
138. 
he 9 
d 
he 5 
d 
he 9 
, an 
he 5 
, an 
he 5 
4.7, 
the 

5 th percentile 
and 
th percentile 
and 
5 th percentile 
8 , and 
th percentile 
0, and 
5 th percentile 

th percentile 

5 th percentile 
d 
th percentile 
d 
th percentile 
and 

95 th percentile 

the length of the Sth 
and 

the length of the Sth 

RULE002 [ANTHRO-RULES] 

If the workforce popula 
Then 1) it is definite 

standing elbow 
2) it is definite 

standing elbow 
3) it is definite 

standing should 
4) it is definite 

standing should 
5) it is definite 

popliteal heigh 
6) it is definite 

popliteal heigh 

tion is FEMALE, 
(100%) that the 95 th percentile 
height is 111.2, and 
(100%) that the Sth percentile 
height is 99.0, and 
(100%) that the 95 th percentile 
er height is 146.6, and 
(100%) that the Sth percentile 
er height is 127.9, and 
(100%) that the 95 th percentile 
t is 43.6, and 
(100%) that the Sth percentile 
t is 37.3, and 

59 



60 

7) it is definite (100%) that the 95 th percentile 
sitting elbow height is 21.k, and 

8) it is definite (100%) that the Sth percentile 
sitting elbow height is 18.8, and 

9) it is definite (100%) that the Sth percentile 
sitting shoulder height is 52.2, and 

10) it is definite (100%) that the 95 th percentile 
body depth is 27.6, and 

11) it is definite (100%) that the length of the Sth 
percentile forearm is 32.5, and 

12) it is definite (100%) that the length of the Sth 
percentile arm is 60.2, ' 

RULE003 [ANTHRO-RULES] 

If the workf 
Then 1) it i 

Stan 
it i 
Stan 

I 

it i 
Stan 
lit i 

/̂  

v-" 

V 

4 

5 

6 

7 

8 

9 

10) 

12) 

Stan 
it i 
popl 
it i 
popl 
it i 
sitt 
It 1 
sitt 
it i 
sitt 
it 
bod 
it 
per 
it 
per 

orce 
s de 
ding 
s de 
ding 
s de 
ding 
s de 
ding 
s de 
itea 
s de 
itea 
s de 
ing 
s de 
ing 
s de 
ing 
is d 
y de 
is d 
cent 
is d 
cent 

pop 
f ini 
elb 

f ini 
elb 

f ini 
sho 

f ini 
sho 

f ini 
1 he 
f ini 
1 he 
f ini 
elbo 
f ini 
elbo 
f ini 
shou 
ef in 
pth 
ef in 
ile 
ef in 
ile 

ulati 
te (1 

ow 
te 
ow 
te 

he 
(1 
he 
(1 

ulder 
te (1 
ulder 
te (1 
ight 
te (1 
ight 
te (1 
w hei 
te (1 
w hei 
te (1 
Ider 
ite ( 
is 34 
ite ( 
f orea 
ite ( 
arm i 

on 1 
00%) 
ight 
00%) 
ight 
00%) 
hei 

00%) 
hei 

00%) 
is 4 
00%) 
is 3 
00%) 
ght 
00%) 
ght 
00%) 
heig 
100% 
.0, 
100% 
rm i 
100% 
s 60 

s MI 
tha 
is 
tha 
is 
tha 

ght 
tha 

ght 
tha 

8.7, 
tha 

7.3, 
tha 

is 2 
tha 

is 1 
tha 

ht i 
) th 
and 
) th 
s 32 
) th 
.2. 

XED, 
t the 9 
120.4, 
t the 5 
99.0, a 
t the 9 
is 156. 
t the 5 
is 127. 
t the 9 
and 

t the 
and 

t the 9 
4 , an 
the 5 
8 , an 
the 5 
52.2, 
the 

7. 
t 
8. 
t 
s 
at 

5 th percentile 
and 
th percentile 
nd 
5 th percentile 
8 , and 
th percentile 
9, and 
5 th percentile 

Sth percentile 

5 th percentile 
d 
th percentile 
d 
th percentile 
and 

95 th percentile 

at the length of the Sth 
.5, and 
at the length of the Sth 

RULE043 [ANTHRO-RULES] 

If the workforce population is MALE, 
Then 1) it is definite (100%) that the Sth percentile thigh 

width is 14.2, and 
2) it is definite (100%) that the 95 th percentile 

thigh width is 18.5. 



61 

RULE044 [ANTHRO-RULES] 

If the workforce population is FEMALE, 
Then 1) it is definite (100%) that the Sth percentile thigh 

width is 12.4, and 
2) it is definite (100%) that the 95 th percentile 

thigh width is 16.S. 

RULE04S [ANTHRO-RULES] 

If the workforce population is MIXED, 
Then 1) it is definite (100%) that the Sth percentile thigh 

width is 12,4, and 
2) it is definite (100%) that the 95 th percentile 

thigh width is 18,5, 

Rule Group DATA-CHECKING-RULES 

RULE049 [DATA-CHECKING-RULES] 

If 1 

2 

the Sth percentile standing shoulder height is less 
than 138, or 
the 95 th percentile standing shoulder height is 
greater than 156,8, or 

3) the Sth percentile sitting shoulder height is less 
than 54 , 7 , or 

4) the 95 th percentile body depth is greater than 34, 
or 

5) the Sth percentile thigh width is less than 14.2, or 
6) the 95 th percentile thigh width is greater than 

18,5, or 
7) the length of the Sth percentile forearm is less than 

37.1, or 
8) the length of the Sth percentile arm is less than 

68.3, or 
9) the Sth percentile standing elbow height is less than 

105.6, or 
10) the 95 th percentile standing elbow height is 

greater than 120.4, or 
11) the Sth percentile sitting elbow height is less than 

18.8, or 
12) the 95 th percentile sitting elbow height is greater 

than 27.4, or 
13) the Sth percentile popliteal height is less than 

42.4, or 
14) the 95 th percentile popliteal height is greater 

than 48.7, 
Then it is definite (100%) that there is a probable error 

in the male anthropometric data. 



62 

RULEOSO [DATA-CHECKING-RULES] 

If 1 

2 

3 

4 

5 
6 

7 

8 

9 

the Sth percentile standing shoulder height is less 
than 127.9, or 
the 95 th percentile standing shoulder height is 
greater than 146.6, or 
the Sth percentile sitting shoulder height is less 
than 52.2, or 
the 95 th percentile body depth is greater than 27.6, 
or 
the Sth percentile thigh width is less than 12.4, or 
the 95 th percentile thigh width is greater than 
16.5, or 
the length of the Sth percentile forearm is less than 
32.5, or 
the length of the Sth percentile arm is less than 
60.2, or 
the Sth percentile standing elbow height is less than 
99, or 

10) the 95 th percentile standing elbow height is 
greater than 111.2, or 

11) the Sth percentile sitting elbow height is less than 
18.8, or 

12) the 95 th percentile sitting elbow height is greater 
than 27.4, or 

13) the Sth percentile popliteal height is less than 
37.3, or 

14) the 95 th percentile popliteal height is greater 
than 43.6, 

Then it is definite (100%) that there is a probable error 
in the female anthropometric data. 

RULEOSl [DATA-CHECKING-RULES] 

If 1 

2 

3 

4 

5 
6 

7 

8 

9 

the Sth percentile standing shoulder height is less 
than 127.9, or 
the 95 th percentile standing shoulder height is 
greater than 156.8, or 
the Sth percentile sitting shoulder height is less 
than 52.2, or 
the 95 th percentile body depth is greater than 34, 
or 
the Sth percentile thigh width is less than 12.4, or 
the 95 th percentile thigh width is greater than 
18.5, or 
the length of the Sth percentile forearm is less than 
32.5, or 
the length of the Sth percentile arm is less than 
60.2, or 
the Sth percentile standing elbow height is less than 
99, or 



63 

10) the 95 th percentile standing elbow height is 
greater than 120.4, or 

11) the Sth percentile sitting elbow height is less than 
18.8, or 

12) the 95 th percentile sitting elbow height is greater 
than 27.4, or 

13) the Sth percentile popliteal height is less than 
37.3, or 

14) the 95 th percentile popliteal height is greater 
than 48.7, 

Then it is definite (100%) that there is a probable error 
in the mixed anthropometric data. 

RULE0S2 [DATA-CHECKING-RULES] 

If 1) the workforce population is SPECIFIC, and 
2) the sex of your specific population is MALE, and 
3) there is a probable error in the male anthropometric 

data, 
Then it is definite (100%) that DATA CHECKING is The 

following conclusions were made by THE ERGONOMIST 
concerning the current workplace design. However, a 
probable error was detected in the anthropometric data 
that you entered. Please re-check your data and run 
another consultation.. 

RULE053 [DATA-CHECKING-RULES] 

If 1) the workforce population is SPECIFIC, and 
2) the sex of your specific population is FEMALE, and 
3) there is a probable error in the female 

anthropometric data. 
Then it is definite (100%) that DATA CHECKING is The 

following conclusions were made by THE ERGONOMIST 
concerning the current workplace design. However, a 
probable error was detected in the anthropometric data 
that you entered. Please re-check your data and run 
another consultation.. 

RULE0S4 [DATA-CHECKING-RULES] 

If 1) the workforce population is SPECIFIC, and 
2) the sex of your specific population is MIXED, and 
3) there is a probable error in the mixed anthropometric 

data , 
Then it is definite (100%) that DATA CHECKING is The 

following conclusions were made by THE ERGONOMIST 
concerning the current workplace design. However, a 
probable error was detected in the anthropometric data 



64 

that you entered. Please re-check your data and run 
another consultation.. 

RULEOSS [DATA-CHECKING-RULES] 

If 1) the workforce population is MALE, or 
2) the workforce population is FEMALE, or 
3) the workforce population is MIXED, or 
4) there is not a probable error in the male 

anthropometric data, or 
5) there is not a probable error in the female 

anthropometric data, or 
6) there is not a probable error in the mixed 

anthropometric data. 
Then it is definite (100%) that DATA CHECKING is The 

following conclusions were made by THE ERGONOMIST 
concerning the current worplace design,. 

Rule Group DIMENSIONING-RULES 

RULE020 [DIMENSIONING-RULES] 

If the classification of the task is PRECISION-WORK, 
Then 1) it is definite (100%) that the work classification 

correction factor-1 is 12, and 
2) it is definite (100%) that the work classification 

correction factor-2 is 8. 

RULE021 [DIMENSIONING-RULES] 

If the classification of the task is LIGHT-WORK, 
Then 1) it is definite (100%) that the work classification 

correction factor-1 is 5, and 
2)\it is definite (100%) that the work classification 

correction factor-2 is -5. 

RULE022 [DIMENSIONING-RULES] 

If the classification of the task is HEAVY-WORK, 
Then 1) it is definite (100%) that the work classification 

correction factor-1 is 0, and 
2) it is definite (100%) that the work classification 

correction factor-2 is -13. 



65 

RULE023 [DIMENSIONING-RULES] 

If the classification of task is VDT/KEYBOARD-OPERATION, 
Then 1) it is definite (100%) that the work classification 

correction factor-1 is 0, and 
2) it is definite (100%) that the work classification 

correction factor-2 is -5. 

RULE024 [DIMENSIONING-RULES] 

If 1) the workstation type is SIT, and 
2) the design type is FULLY-ADJUSTABLE, 

Then 1) it is definite (100%) that work surface upper ht. 
in centimeters is [ [ the 95 th percentile 
popliteal height plus the 95 th percentile sitting 
elbow height ] plus the work classification 
correction factor-1 ], and 

2) it is definite (100%) that work surface lower ht. 
in centimeters is [ [ the Sth percentile popliteal 
height plus the Sth percentile sitting elbow 
height ] plus the work classification correction 
factor-1 ], and 

3) it is definite (100%) that chair upper height in 
centimeters is [ the 95 th percentile popliteal 
height minus 0 ], and 

4) it is definite (100%) that chair lower height in 
centimeters is [ the Sth percentile popliteal 
height minus 0 ], and 

5) it is definite (100%) that footrest upper height in 
centimeters is [ the 95 th percentile popliteal 
height minus the Sth percentile popliteal height 
] , and 

6) it is definite (100%) that footrest lower height in 
centimeters is 0, and 

7) it is definite (100%) that normal vertical reach in 
centimeters is [ [ the Sth percentile popliteal 
height plus the 5th percentile sitting elbow 
height ] plus the length of the Sth percentile 
forearm ], and 

8) it is definite (100%) that maximum vertical reach 
in centimeters is [ [ the Sth percentile popliteal 
height plus the Sth percentile sitting shoulder 
height ] plus the length of the Sth percentile arm 
] , and 

9) it is definite (100%) that normal horizontal reach 
in centimeters is [ the length of the Sth 
percentile forearm minus [ .5 times the 95 th 
percentile body depth ] ], and 

10) it is definite (100%) that maximum horizontal 
reach in centimeters is [ the length of the Sth 
percentile arm minus [ .5 times the 95 th 



66 

percentile body depth ] ]. 

RULE025 [DIMENSIONING-RULES] 

If 1) the workstation type is STAND, and 
2) the design type is FULLY-ADJUSTABLE, 

Then 1) it is definite (100%) that work surface upper ht. 
in centimeters is [ the 95 th percentile standing 
elbow height plus the work classification 
correction factor-2 ] , and 

2) it is definite (100%) that work surface lower ht. 
in centimeters is [ the Sth percentile standing 
elbow height plus the work classification 
correction factor-2 ], and 

3) it is definite (100%) that chair upper height in 
centimeters is NONE, and 

4) it is definite (100%) that chair lower height in 
centimeters is NONE, and 

5) it is definite (100%) that footrest upper height in 
centimeters is NONE, and 

6) it is definite (100%) that footrest lower height in 
^-^ centimeters is NONE, and 
7) it is definite (100%) that normal vertical reach in 

centimeters is [ the Sth percentile standing elbow 
height plus the length of the Sth'^percentile 
forearm ], and 

8) it is definite (100%) that maximum vertical reach 
-̂  in centimeters is [ the Sth percentile standing 

shoulder height plus the length of the Sth 
percentile arm ], and 

9) it is definite (100%) that normal horizontal reach 
. -yln centimeters is [ the length of the Sth 

percentile forearm minus [ .5 times the 95 th 
percentile body depth ] ], and ^ 

10) it is definite (100%) that maximum horizontal 
^ ̂  )reach in centimeters is [ the length of the Sth 

percentile arm minus [ .5 times the 95 th 
percentile body depth ] ] . 

RULE026 [DIMENSIONING-RULES] 

If 1) the workstation type is SIT/STAND, and 
2) the design type is FULLY-ADJUSTABLE, 

Then 1) it is definite (100%) that work surface upper ht. 
in centimeters is [ the 95 th percentile standing 
elbow height plus the work classification 
correction factor-2 ] , and 

2) it is definite (100%) that work surface lower ht. 
in centimeters is [ the Sth percentile standing 
elbow height plus the work classification 



67 

correction factor-2 ], and 
3) it is definite (100%) that chair upper height in 

centimeters is [ the 95 th percentile standing 
elbow height minus the Sth percentile sitting 
elbow height J, and 

4) it is definite (100%) that chair lower height in 
centimeters is [ the Sth percentile standing elbow 
height minus the 95 th percentile sitting elbow 
height ] , and 

5) it is definite (100%) that footrest upper height in 
centimeters is [ [ the 95 th percentile standing 
elbow height minus the Sth percentile sitting 
elbow height ] minus the Sth percentile popliteal 
height ] , and 

6) it is definite (100%) that footrest lower height in 
centimeters is 0, and 

7) it is definite (100%) that normal vertical reach in 
centimeters is [ the Sth percentile standing elbow 
height plus the length of the Sth percentile 
forearm ], and 

8) it is definite (100%) that maximum vertical reach 
in centimeters is [ the Sth percentile standing 
shoulder height plus the length of the Sth 
percentile arm ] , and 

9) it is definite (100%) that normal horizontal reach 
in centimeters is [ the length of the Sth 
percentile forearm minus [ .5 times the 95 th 
percentile body depth ] ], and 

10) it is definite (100%) that maximum horizontal 
reach in centimeters is [ the length of the Sth 
percentile arm minus the 95 th percentile body 
depth ] . 

RULE027 [DIMENSIONING-RULES] 

If 1) the workstation type is SIT, and 
2) the design type is FIXED-WORK-SURFACE, 

Then 1) it is definite (100%) that work surface upper ht. 
in centimeters is [ [ the 95 th percentile 
popliteal height plus the 95 th percentile sitting 
elbow height ] plus the work classification 
correction factor-1 ] , and 

2) it is definite (100%) that work surface lower ht. 
in centimeters is NONE, and 

3) it is definite (100%) that chair upper height in 
centimeters is [ [ the 95 th percentile popliteal 
height plus the 95 th percentile sitting elbow 
height ] minus the Sth percentile sitting elbow 
height ], and 

4) it is definite (100%) that chair lower height in 
centimetjers is [ the 95 th percentile popliteal 



68 

height minus 0 ], and 
5) it is definite (100%) that footrest upper height in 

centimeters is [ [ [ the 95 th percentile popliteal 
height plus the 95 th percentile sitting elbow 
height ] minus the Sth percentile sitting elbow 
height ] minus the Sth percentile popliteal height 
] , and 

6) it is definite (100%) that footrest lower height in 
centimeters is 0, and 

7) it is definite (100%) that normal vertical reach in 
centimeters is [ [ the 95 th percentile popliteal 
height plus the 95 th percentile sitting elbow 
height ] plus the length of the Sth percentile 
forearm ], and 

8) it is definite (100%) that maximum vertical reach 
in centimeters is [ [ the 95 th percentile 
popliteal height plus the Sth percentile sitting 
shoulder height ] plus the length of the Sth 
percentile arm ], and 

9) it is definite (100%) that normal horizontal reach 
in centimeters is [ the length of the Sth 
percentile forearm minus [ .5 times the 95 th 
percentile body depth ] ] , and 

10) it is definite (100%) that maximum horizontal 
reach in centimeters is [ the length of the Sth 
percentile arm minus [ .5 times the 95 th 
percentile body depth ] ]. 

RULE028 [DIMENSIONING-RULES] 

If 1) the workstation type is STAND, and 
2) the design type is FIXED-WORK-SURFACE, 

Then 1) it is definite (100%) that work surface upper ht. 
in centimeters is [ the 95 th percentile standing 
elbow height plus the work classification 
correction factor-2 ], and 

2) it is definite (100%) that work surface lower ht. 
in centimeters is NONE, and 

3) it is definite (100%) that chair upper height in 
centimeters is NONE, and 

4) it is definite (100%) that chair lower height in 
centimeters is NONE, and 

5) it is definite (100%) that footrest upper height in 
centimeters is [ the 95 th percentile standing 
elbow height minus the Sth percentile standing 
elbow height ] , and 
it is definite (100%) that footrest lower height in 
centimeters is 0, and 
it is definite (100%) that normal vertical reach in 
centimeters is [ the 95 th percentile standing 
elbow height plus the length of the Sth percentile 



69 

forearm ], and 
8)iit is definite (100%) that maximum vertical reach 

in centimeters is [ the Sth percentile standing 
shoulder height plus the length of the Sth 

^̂v̂̂js percentile arm ] , and 
9) it is definite (100%) that normal horizontal reach 
\,\in centimeters is [ the length of the Sth 

'percentile forearm minus [ .5 times the 95 th 
^percentile body depth ] ], and 

10) it is definite (100%) that maximum horizontal 
reach in centimeters is [ the length of the Sth 

^ percentile arm minus [ .5 times the 95 th 
percentile body depth ] ] . 

RULE029 [DIMENSIONING-RULES] 

If 1) the workstation type is SIT/STAND, and 
2) the design type is FIXED-WORK-SURFACE, 

Then 1) it is definite (100%) that work surface upper ht. 
in centimeters is [ the 95 th percentile standing 
elbow height plus the work classification 
correction factor-2 J, and 

2) it is definite (100%) that work surface lower ht. 
in centimeters is NONE, and 

3) it is definite (100%) that chair upper height in 
centimeters is [ the 95 th percentile standing 
elbow height minus the Sth percentile sitting 
elbow height ], and 

4) it is definite (100%) that chair lower height in 
centimeters is [ the 95 th percentile standing 
elbow height minus the 95 th percentile sitting 
elbow height ], and 

5) it is definite (100%) that footrest upper height in 
centimeters is [ [ the 95 th percentile standing 
elbow height minus the Sth percentile sitting 
elbow height ] minus the Sth percentile popliteal 
height ], and 

6) it is definite (100%) that footrest lower height in 
centimeters is 0, and 

7) it is definite (100%) that normal vertical reach in 
centimeters is [ the 95 th percentile standing 
elbow height plus the length of the Sth percentile 
forearm ], and 

8) it is definite (100%) that maximum vertical reach 
in centimeters is [ [ [ the 95 th percentile 
standing elbow height minus the 95 th percentile 
sitting elbow height ] plus the Sth percentile 
sitting shoulder height ] plus the length of the 
Sth percentile arm ], and 

9) it is definite (100%) that normal horizontal reach 
in centimeters is [ the length of the Sth 



70 

percentile forearm minus [ .5 times the 95 th 
percentile body depth ] ], and 

10) it is definite (100%) that maximum horizontal 
reach in centimeters is [ the length of the Sth 
percentile arm minus [ .5 times the 95 th 
percentile body depth ] ]. 

RULE030 [DIMENSIONING-RULES] 

If 1) the workstation type is SIT, and 
2) the design type is FIXED-CHAIR, 

Then 1) it is definite (100%) that work surface upper ht. 
in centimeters is [ [ the 95 th percentile 
popliteal height plus the 95 th percentile sitting 
elbow height ] plus the work classification 
correction factor-1 ], and 

2) it is definite (100%) that work surface lower ht. 
in centimeters is [ [ the 95 th percentile 
popliteal height plus the Sth percentile sitting 
elbow height ] plus the work classification 
correction factor-1 ], and 

3) it is definite (100%) that chair upper height in 
centimeters is [ the 95 th percentile popliteal 
height minus 0 ], and 

4) it is definite (100%) that chair lower height in 
centimeters is NONE, and 

5) it is definite (100%) that footrest upper height in 
centimeters is [ the 95 th percentile popliteal 
height minus the Sth percentile popliteal height 
] , and 

6) it is definite (100%) that normal vertical reach in 
centimeters is [ [ the 95 th percentile popliteal 
height plus the Sth percentile sitting elbow 
height ] plus the length of the Sth percentile 
forearm ], and 

7)- it is definite (100%) that maximum vertical reach 
in centimeters is [ [ the 95 th percentile 
popliteal height plus the Sth percentile sitting 
shoulder height ] plus the length of the Sth 
percentile arm ], and 

8) it is definite (100%) that normal horizontal reach 
in centimeters is [ the length of the Sth 
percentile forearm minus [ .5 times the 95 th 
percentile body depth ] ], and 

9) it is definite (100%) that maximum horizontal reach 
in centimeters is [ the length of the Sth 
percentile arm minus [ .5 times the 95 th 
percentile body depth ] ] . 



71 

RULE031 [DIMENSIONING-RULES] 

If 1) the workstation type is STAND, and 
2) the design type is FIXED-CHAIR, 

Then 1) it is definite (100%) that work surface upper ht. 
in centimeters is [ the 95 th percentile standing 
elbow height plus the work classification 
correction factor-2 ], and 

2) it is definite (100%) that work surface lower ht. 
in centimeters is [ the Sth percentile standing 
elbow height plus the work classification 
correction factor-2 ], and 

3) it is definite (100%) that chair upper height in 
centimeters is NONE, and 

4) it is definite (100%) that chair lower height in 
centimeters is NONE, and 

5) it is definite (100%) that footrest upper height in 
centimeters is NONE, and 

6) it is definite (100%) that footrest lower height in 
centimeters is NONE, and 

7) it is definite (100%) that normal vertical reach in 
centimeters is [ the Sth percentile standing elbow 
height plus the length of the Sth percentile 
forearm ], and 

8) it is definite (100%) that maximum vertical reach 
in centimeters is [ the Sth percentile standing 
shoulder height plus the length of the Sth 
percentile arm ], and 

9) it is definite (100%) that normal horizontal reach 
in centimeters is [ the length of the Sth 
percentile forearm minus [ .5 times the 95 th 
percentile body depth ] ], and 

10) it is definite (100%) that maximum horizontal 
reach in centimeters is [ the length of the Sth 
percentile arm minus [ .5 times the 95 th 
percentile body depth ] ]. 

RULE032 [DIMENSIONING-RULES] 

If 1) the workstation type is SIT/STAND, and 
2) the design type is FIXED-CHAIR, 

Then 1) it is definite (100%) that work surface upper ht. 
in centimeters is [ [ [ the 95 th percentile 
standing elbow height minus the Sth percentile 
sitting elbow height ] plus the 95 th percentile 
sitting elbow height ] plus the work 
classification correction factor-2 ], and 

2) it is definite (100%) that work surface lower ht. 
in centimeters is [ the 95 th percentile standing 
elbow height plus the work classification 
correction factor-2 ], and 



72 

3) it is definite (100%) that chair upper height in 
centimeters is [ the 95 th percentile standing 
elbow height minus the Sth percentile sitting 
elbow height ], and 

4) it is definite (100%) that chair lower height in 
centimeters is NONE, and 

5) it is definite (100%) that footrest upper height in 
centimeters is [ [ the 95 th percentile standing 
elbow height minus the Sth percentile sitting 
elbow height ] minus the Sth percentile popliteal 
height ], and 

6) it is definite (100%) that footrest lower height in 
centimeters is 0, and 

7) it is definite (100%) that normal vertical reach in 
centimeters is [ the Sth percentile standing elbow 
height plus the length of the Sth percentile 
forearm ], and 

8) it is definite (100%) that maximum vertical reach 
in centimeters is [ the Sth percentile standing 
shoulder height plus the length of the Sth 
percentile arm ], and 

9) it is definite (100%) that normal horizontal reach 
in centimeters is [ the length of the Sth 
percentile forearm minus [ .5 times the 95 th 
percentile body depth ] ], and 

10) it is definite (100%) that maximum horizontal 
reach in centimeters is [ the length of the Sth 
percentile arm minus [ .5 times the 95 th 
percentile body depth ] ]. 

RULE046 [DIMENSIONING-RULES] 

If 1) the workstation type is SIT, and 
2) the design type is STATIONARY, 

Then 1) it is definite (100%) that work surface upper ht. 
in centimeters is [ [ the 95 th percentile 
popliteal height plus the 95 th percentile sitting 
elbow height ] plus the work classification 
correction factor-1 ], and 

2) it is definite (100%) that work surface lower ht. 
in centimeters is NONE, and 

3) it is definite (100%) that chair upper height in 
centimeters is [ [ the 95 th percentile popliteal 
height plus the 95 th percentile sitting elbow 
height ] minus [ [ the Sth percentile sitting 
elbow height plus the 95 th percentile sitting 
elbow height ] divided by 2 ] ], and 

4) it is definite (100%) that chair lower height in 
centimeters is NONE, and 

5) it is definite (100%) that footrest upper height in 
centimeters is [ chair upper height in centimeters 



73 

minus [ [ the Sth percentile popliteal height plus 
the 95 th percentile popliteal height ] divided by 
2 ] ], and 

6) it is definite (100%) that footrest lower height in 
centimeters is NONE, and 

7) it is definite (100%) that normal vertical reach in 
centimeters is [ [ chair upper height in 
centimeters plus the Sth percentile sitting elbow 
height ] plus the length of the Sth percentile 
forearm ], and 

8) it is definite (100%) that maximum vertical reach 
in centimeters is [ [ chair upper height in 
centimeters plus the Sth percentile sitting 
shoulder height ] plus the length of the Sth 
percentile arm ], and 

9) it is definite (100%) that normal horizontal reach 
in centimeters is [ the length of the Sth 
percentile forearm minus [ the 95 th percentile 
body depth divided by 2 ] ], and 

10) it is definite (100%) that maximum horizontal 
reach in centimeters is [ the length of the Sth 
percentile arm minus [ the 95 th percentile body 
depth divided by 2 ] ]. 

RULE047 [DIMENSIONING-RULES] 

If 1) the workstation type is STAND, and 
2) the design type is STATIONARY, 

Then 1) it is definite (100%) that work surface upper ht. 
in centimeters is [ the 95 th percentile standing 
elbow height plus the work classification 
correction factor-2 ], and 

2) it is definite (100%) that work surface lower ht. 
in centimeters is NONE, and 

3) it is definite (100%) that chair upper height in 
centimeters is NONE, and 

4) it is definite (100%) that chair lower height in 
centimeters is NONE, and 

5) it is definite (100%) that footrest upper height in 
centimeters is [ the 95 th percentile standing 
elbow height minus [ [ the Sth percentile standing 
elbow height plus the 95 th percentile standing 
elbow height ] divided by 2 ] ] , and 

6) it is definite (100%) that footrest lower height in 
centimeters is NONE, and 

7) it is definite (100%) that normal vertical reach in 
centimeters is [ [ footrest upper height in 
centimeters plus the Sth percentile standing elbow 
height ] plus the length of the Sth percentile 
forearm ], and 

8) it is definite (100%) that maximum vertical reach 



74 

in centimeters is [ [ footrest upper height in 
centimeters plus the Sth percentile standing 
shoulder height ] plus the length of the Sth 
percentile arm ], and 

9) it is definite (100%) that normal horizontal reach 
in centimeters is [ the length of the Sth 
percentile forearm minus [ the 95 th percentile 
body depth divided by 2 ] ], and 

10) it is definite (100%) that maximum horizontal 
reach in centimeters is [ the length of the Sth 
percentile arm minus [ the 95 th percentile body 
depth divided by 2 ] ] . 

RULE048 [DIMENSIONING-RULES] 

If 1) the workstation type is SIT/STAND, and 
2) the design type is STATIONARY, 

Then 1) it is definite (100%) that work surface upper ht. 
in centimeters is [ the 95 th percentile standing 
elbow height plus the work classification 
correction factor-2 ], and 

2) it is definite (100%) that work surface lower ht. 
in centimeters is NONE, and 

3) it is definite (100%) that chair upper height in 
centimeters is [ the 95 th percentile standing 
elbow height minus [ [ the Sth percentile sitting 
elbow height plus the 95 th percentile sitting 
elbow height ] divided by 2 ] ], and 

4) it is definite (100%) that chair lower height in 
centimeters is NONE, and 

5) it is definite (100%) that footrest upper height in 
centimeters is [ chair upper height in centimeters 
minus [ [ the Sth percentile popliteal height plus 
the 95 th percentile popliteal height ] divided by 
2 ] ] , and 

6) it is definite (100%) that footrest lower height in 
centimeters is NONE, and 

7) it is definite (100%) that normal vertical reach in 
centimeters is [ [ chair upper height in 
centimeters plus the Sth percentile sitting elbow 
height ] plus the length of the Sth percentile 
forearm ], and 

8) it is definite (100%) that maximum vertical reach 
in centimeters is [ [ chair upper height in 
centimeters plus the Sth percentile sitting 
shoulder height ] plus the length of the Sth 
percentile arm ] , and 

9) it is definite (100%) that normal horizontal reach 
in centimeters is [ the length of the Sth 
percentile forearm minus [ the 95 th percentile 
body depth divided by 2 ] ], and 



75 

10) it is definite (100%) that maximum horizontal 
reach in centimeters is [ the length of the Sth 
percentile arm minus [ the 95 th percentile body 
depth divided by 2 ] ]. 

Rule Group STATION-TYPE-RULES 

RULE004 [STATION-TYPE-RULES] 

If 1) the classification of the task is PRECISION-WORK, and 
2) the requirement to use foot controls, and 
3) the requirement to cover a large work area. 

Then it is definite (100%) that the workstation type is 
SIT/STAND. 

RULEOOS [STATION-TYPE-RULES] 

If 1) the classification of the task is PRECISION-WORK, and 
2) the requirement to use foot controls, and 
3) the requirement to cover a large work area is not 

true , 
Then it is definite (100%) that the workstation type is 

SIT. 

RULE006 [STATION-TYPE-RULES] 

If 1) the classification of the task is PRECISION-WORK, and 
2) the requirement to use foot controls is not true, and 
3) the requirement to cover a large work area is not 

true , 
Then it is definite (100%) that the workstation type is 

SIT/STAND. 

RULE007 [STATION-TYPE-RULES] 

If 1) the classification of the task is PRECISION-WORK, and 
2) the requirement to use foot controls is not true, and 
3) the requirement to cover a large work area. 

Then it is definite (100%) that the workstation type is 
SIT/STAND. 

RULE008 [STATION-TYPE-RULES] 

If 1) the classification of the task is LIGHT-WORK, and 
2) the requirement to use foot controls, and 
3) the requirement to cover a large work area. 



76 

Then it is definite (100%) that the workstation type is 
SIT/STAND. 

RULE009 [STATION-TYPE-RULES] 

If 1) the classification of the task is LIGHT-WORK, and 
2) the requirement to use foot controls, and 
3) the requirement to cover a large work area is not 

true , 
Then it is definite (100%) that the workstation type is 

SIT. 

RULEOIO [STATION-TYPE-RULES] 

If 1) the classification of the task is LIGHT-WORK, and 
2) the requirement to use foot controls is not true, 
3) the requirement to cover a large work area is not 

true, 
Then it is definite (100%) that the workstation type is 

SIT/STAND. 

RULEOll [STATION-TYPE-RULES] 

If 1) the classification of the task is HEAVY-WORK, and 
2) the requirement to use foot controls, and 
3) the requirement to cover a large work area. 

Then it is definite (100%) that the workstation type is 
SIT/STAND. 

RULE013 [STATION-TYPE-RULES] 

If 1) the classification of the task is HEAVY-WORK, and 
2) the requirement to use foot controls, and 
3) the requirement to cover a large work area is not 

true , 
Then it is definite (100%) that the workstation type is 

SIT/STAND. 

and 

If 1) the classification of the task is LIGHT-WORK, and 
2) the requirement to use foot controls is not true, and 
3) the requirement to cover a large work area. 

Then it is definite (100%) that the workstation type is 
SIT/STAND. 

RULE012 [STATION-TYPE-RULES] 



77 

RULE014 [STATION-TYPE-RULES] 

If 1) the classification of the task is HEAVY-WORK, and 
2) the requirement to use foot controls is not true, and 
3) the requirement to cover a large work area is not 

true , 
Then it is definite (100%) that the workstation type is 

STAND. 

RULEOIS [STATION-TYPE-RULES] 

If 1) the classification of the task is HEAVY-WORK, and 
2) the requirement to use foot controls is not true, and 
3) the requirement to cover a large work area. 

Then it is definite (100%) that the workstation type is 
STAND. 

RULE016 [STATION-TYPE-RULES] 

If 1) the classification of the task is 
VDT/KEYBOARD-OPERATION, and 

2) the requirement to use foot controls, and 
3) the requirement to cover a large work area. 

Then it is definite (100%) that the workstation type is 
SIT/STAND. 

RULE017 [STATION-TYPE-RULES] 

If 1) the classification of the task is 
VDT/KEYBOARD-OPERATION, and 

2) the requirement to use foot controls, and 
3) the requirement to cover a large work area is not 

true, 
Then it is definite (100%) that the workstation type is 

SIT. 

RULE018 [STATION-TYPE-RULES] 

If 1) the classification of the task is 
VDT/KEYBOARD-OPERATION, and 

2) the requirement to use foot controls is not true, and 
3) the requirement to cover a large work area is not 

true , 
Then it is definite (100%) that the workstation type is 

SIT. 



78 

RULE019 [STATION-TYPE-RULES] 

If 1) the classification of the task is 
VDT/KEYBOARD-OPERATION, and 

2) the requirement to use foot controls is not true, and 
3) the requirement to cover a large work area. 

Then it is definite (100%) that the workstation type is 
SIT/STAND. 

Rule Group THIGH-CLEARANCE-RULES 

RULE033 [THIGH-CLEARANCE-RULES] 

If there is not a thigh clearance problem, 
Then it is definite (100%) that the approximate reduction 

to the population that is to be accommodated at this 
workplace, due to a thigh clearance problem is 
NO-REDUCTION. 

RULE034 [THIGH-CLEARANCE-RULES] 

If there is a thigh clearance problem. 
Then 1) it is definite (100%) that the thigh clearance 

difference is [ the 95 th percentile thigh width 
minus [ [ work surface upper ht. in centimeters 
minus the table thickness ] minus chair upper 
height in centimeters ] ], and 

2) it is definite (100%) that the Sth percentile 
thigh-width / popliteal combination is [ the Sth 
percentile thigh width plus the Sth percentile 
popliteal height ], and 

3) it is definite (100%) that the 95 th percentile 
thigh width / popliteal combination is [ the 95 th 
percentile thigh width plus the 95 th percentile 
popliteal height J, and 

4) it is definite (100%) that the estimated thigh 
width mean is [ [ the Sth percentile thigh-width / 
popliteal combination plus the 95 th percentile 
thigh width / popliteal combination ] divided by 
] , and 

5) it is definite (100%) that the estimated thigh 
width standard deviation is [ [ the 95 th 
percentile thigh width / popliteal combination 
minus the estimated thigh width mean ] divided by 
1.645 ], and 

6) it is definite (100%) that the critical 
thigh-clearance height is [ the 95 th percentile 
thigh width / popliteal combination minus the 
thigh clearance difference ]. 



79 

RULE03S [THIGH-CLEARANCE-RULES] 

If 1) [ [ the critical thigh-clearance height minus the 
estimated thigh width mean ] divided by the 
estimated thigh width standard deviation ] is less 
than or equal to 1.645, and 

2) [ [ the critical thigh-clearance height minus the 
estimated thigh width mean ] divided by the 
estimated thigh width standard deviation ] is greater 
than 1.28, 

Then it is definite (100%) that the approximate reduction 
to the population that is to be accommodated at this 
workplace, due to a thigh clearance problem is <S%. 

RULE036 [THIGH-CLEARANCE-RULES] 

If 1) [ [ the critical thigh-clearance height minus the 
estimated thigh width mean ] divided by the 
estimated thigh width standard deviation ] is less 
than or equal to 1.28, and 

2) [ [ the critical thigh-clearance height minus the 
estimated thigh width mean ] divided by the 
estimated thigh width standard deviation ] is greater 
than 1.035, 

Then it is definite (100%) that the approximate reduction 
to the population that is to be accommodated at this 
workplace, due to a thigh clearance problem is 5-10%. 

RULE037 [THIGH-CLEARANCE-RULES] 

If 1) [ [ the critical thigh-clearance height minus the 
estimated thigh width mean ] divided by the 
estimated thigh width standard deviation ] is less 
than or equal to 1.035, and 

2) [ [ the critical thigh-clearance height minus the 
estimated thigh width mean ] divided by the 
estimated thigh width standard deviation ] is greater 
than .675, 

Then it is definite (100%) that the approximate reduction 
to the population that is to be accommodated at this 
workplace, due to a thigh clearance problem is ilO-
20%1I. 

RULE038 [THIGH-CLEARANCE-RULES] 

If 1) [ [ the critical thigh-clearance height minus the 
estimated thigh width mean ] divided by the 
estimated thigh width standard deviation ] is less 
than or equal to .675, and 



80 

2) [ [ the critical thigh-clearance height minus the 
estimated thigh width mean ] divided by the 
estimated thigh width standard deviation ] is greater 
than .385, 

Then it is definite (100%) that the approximate reduction 
to the population that is to be accommodated at this 
workplace, due to a thigh clearance problem is i20-
30%^. 

RULE039 [THIGH-CLEARANCE-RULES] 

If 1) [ [ the critical thigh-clearance height minus the 
estimated thigh width mean ] divided by the 
estimated thigh width standard deviation ] is less 
than or equal to ,385, and 

2) [ [ the critical thigh-clearance height minus the 
estimated thigh width mean ] divided by the 
estimated thigh width standard deviation ] is greater 
than .125, 

Then it is definite (100%) that the approximate reduction 
to the population that is to be accommodated at this 
workplace, due to a thigh clearance problem is i30-
40%1I. 

RULE040 [THIGH-CLEARANCE-RULES] 

If 1) [ [ the critical thigh-clearance height minus the 
estimated thigh width mean ] divided by the 
estimated thigh width standard deviation ] is less 
than or equal to .125, and 

2) [ [ the critical thigh-clearance height minus the 
estimated thigh width mean ] divided by the 
estimated thigh width standard deviation ] is greater 
than 0, 

Then it is definite (100%) that the approximate reduction 
to the population that is to be accommodated at this 
workplace, due to a thigh clearance problem is i40-
50%1I. 

RULE041 [THIGH-CLEARANCE-RULES] 

If [ [ the critical thigh-clearance height minus the 
estimated thigh width mean ] divided by the estimated 
thigh width standard deviation ] is less than or equal 
to 0, 

Then it is definite (100%) that the approximate reduction 
to the population that is to be accommodated at this 
workplace, due to a thigh clearance problem is >50%. 



81 

RULE042 [THIGH-CLEARANCE-RULES] 

If 1) the design type is FULLY-ADJUSTABLE, or 
2) the design type is FIXED-CHAIR, or 
3) the workstation type is SIT, or 
4) the workstation type is STAND, 

Then it is definite (100%) that the approximate reduction 
to the population that is to be accommodated at this 
workplace, due to a thigh clearance problem is 
NO-REDUCTION. 

RULE0S6 [THIGH-CLEARANCE-RULES] 

If 1) the design type is STATIONARY, and 
2) the 95 th percentile thigh width is greater than [ [ 

work surface upper ht. in centimeters minus the 
table thickness ] minus chair upper height in 
centimeters ], 

Then it is definite (100%) that there is a thigh clearance 
problem. 

RULE057 [THIGH-CLEARANCE-RULES] 

If 1) the design type is FIXED-WORK-SURFACE, and 
2) the workstation type is SIT/STAND, and 
3) the 95 th percentile thigh width is greater than [ [ 

work surface upper ht. in centimeters minus the 
table thickness ] minus chair upper height in 
centimeters ], 

Then it is definite (100%) that there is a thigh clearance 
problem. 

Parameter Group CONTEXTTYPES 

DIMENSIONAL-DESIGN [CONTEXTTYPES] 

PROMPTEVER: (THE ERGONOMIST is a knowledge-based workplace 
design program developed to provide sound 
ergonomic advice to the industrial 
workstation designer. :line :line :line The 
current objective is to: :line (1) Decide 
what type of workstation to design and :line 
(2) Provide critical dimensions required to 
construct the workstation and :line (3) 
Determine if there is a reduction in the 
accommodated population due to physical 
constraints.) 

PRINTID: WORKPLACE-



82 

PARMGROUP: DIMENSIONAL-DESIGN-PARMS 
RULETYPES: (DATA-CHECKING-RULES THIGH-CLEARANCE-RULES 

DIMENSIONING-RULES STATION-TYPE-RULES 
ANTHRO-RULES) 

INITIALDATA: (WORK-CLASS POPULATION) 
GOALS: (HEADER STATION-TYPE WSU WSL CHU CHL FRU FRL NVR 

MVR NHR MHR POPULATION-REDUCTION) 
DISPLAYRESULTS: T 
UNIQUE: T 

Parameter Group DIMENSIONAL-DESIGN-PARMS 

ARMS [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (the length of the Sth percentile arm) 
PROMPT: (For your specific population, enter the SMALLEST 

Arm Length in centimeters.) 
REPROMPT: (The Arm Length - is the horizontal distance 

from the posterior surface of the shoulder to 
the tip of the extended middle finger.) 

EXPECT: POSNUMB 
CONTAINED-IN: (RULE024 RULE025 RULE026 RULE027 RULE028 

RULE029 RULE030 RULE031 RULE032 RULE046 
RULE047 RULE048) 

USED-BY: (RULE049 RULEOSO RULEOSl) 
UPDATED-BY: (RULEOOl RULE002 RULE003) 

B0DY-DEPTH9S [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (the 95 th percentile body depth) 
PROMPT: (For your specific population, enter the LARGEST 

Body Depth Measurement in centimeters.) 
REPROMPT: (The Body Depth Measurement - is the maximum 

horizontal distance between the vertical planes 
passing through the most anterior and posterior 
on the trunk.) 

EXPECT: POSNUMB 
CONTAINED-IN: (RULE024 RULE025 RULE026 RULE027 RULE028 

RULE029 RULE030 RULE031 RULE032 RULE046 
RULE047 RULE048) 

USED-BY: (RULE049 RULEOSO RULEOSl) 
UPDATED-BY: (RULEOOl RULE002 RULE003) 

CI [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (the work classification correction factor-1) 
EXPECT: NUMB 
CONTAINED-IN: (RULE024 RULE027 RULE030 RULE046) 
UPDATED-BY: (RULE020 RULE021 RULE022 RULE023) 



83 

C2 [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (the work classification correction factor-2) 
EXPECT: NUMB 
CONTAINED-IN: (RULE02S RULE026 RULE028 RULE029 RULE031 

RULE032 RULE047 RULE048) 
UPDATED-BY: (RULE020 RULE021 RULE022 RULE023) 

CHL [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (chair lower height in centimeters) 
EXPECT: POSNUMB 
DICTIONARY: INTERNAL 
CONTAINED-IN: NIL 
UPDATED-BY: (RULE024 RULE02S RULE026 RULE027 RULE028 

RULE029 RULE030 RULE031 RULE032 RULE046 
RULE047 RULE048) 

CHU [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (chair upper height in centimeters) 
EXPECT: POSNUMB 
DICTIONARY: INTERNAL 
CONTAINED-IN: (RULE034) 
USED-BY: (RULE056 RULE057) 
UPDATED-BY: (RULE024 RULE02S RULE026 RULE027 

RULE029 RULE030 RULE031 RULE032 
RULE047 RULE048) 

RULE028 
RULE046 

DESIGN-TYPE [DIMENSIONAL-DESIGN-PARMS] 

TRANS: 
PROMPT 

ASKFIRST 
EXPECT: 

USED-BY: 

(the design type) 
(What type of des 
:line :line FULL 
worksurface, cha 
SURFACE - with a 
:line FIXED CHAI 
and footrest :li 
adjustable featu 

: T 
(FULLY-ADJUSTABLE 
STATIONARY) 
(RULE024 RULE025 
RULE030 RULE031 
RULE048 RULE042 

ign do you wish to construct? 
Y ADJUSTABLE - with adjustable 
ir, and footrest :line FIXED WORK 
djustable chair and footrest 
R - with adjustable worksurface 
ne STATIONARY - with no 
res ) 

FIXED-WORK-SURFACE FIXED-CHAIR 

RULE026 RULE027 
RULE032 RULE046 
RULE0S7) 

RULE028 
RULE047 

RULE029 
RULE056 



84 

ELBOWS [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (the Sth percentile standing elbow height) 
PROMPT: (For your specific population, enter the SMALLEST 

Standing Elbow Height in centimeters.) 
REPROMPT: (The Standing Elbow Height - is the vertical 

distance from the floor to the depression at 
the elbow between the bones of the upper arm 
and forearm,) 

EXPECT: POSNUMB 
CONTAINED-IN: (RULE025 RULE026 RULE028 RULE031 RULE032 

RULE047) 
USED-BY: (RULE049 RULEOSO RULEOSl) 
UPDATED-BY: (RULEOOl RULE002 RULE003) 

ELB0W9S [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (the 95 th percentile standing elbow height) 
PROMPT: (For your specific population, enter the LARGEST 

Standing Elbow Height in centimeters,) 
REPROMPT: (The Standing Elbow Height - is the vertical 

distance from the floor to the depression at 
the elbow between the bones of the upper arm 
and forearm.) 

EXPECT: POSNUMB 
CONTAINED-IN: (RULE025 RULE026 RULE028 RULE029 RULE031 

RULE032 RULE047 RULE048) 
USED-BY: (RULE049 RULEOSO RULEOSl) 
UPDATED-BY: (RULEOOl RULE002 RULE003) 

FEMALE-DATA-ERROR [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (there is a probable error in the female 
anthropometric data) 

USED-BY: (RULE053 RULEOSS) 
UPDATED-BY: (RULEOSO) 

FOOT-CONTROLS [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (the requirement to use foot controls) 
PROMPT: (For this task, is the operator required to use 

any foot controls?) 
REPROMPT: (Is it necessary for the operator to operate 

footpedals, foot switches, or otherwise use his 
feet at the workplace?) 

ASKFIRST: T 
USED-BY: (RULE004 RULEOOS RULE006 RULE007 RULE008 RULE009 

RULEOIO RULEOll RULE012 RULE013 RULE014 RULEOIS 
RULE016 RULE017 RULE018 RULE019) 



85 

FOREARMS [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (the length of the Sth percentile forearm) 
PROMPT: (For your specific population, enter the SMALLEST 

Forearm Length in centimeters.) 
REPROMPT: (The Forearm Length - is the horizontal distance 

from the tip of the elbow to the tip of the 
longest finger.) 

EXPECT: POSNUMB 
CONTAINED-IN: (RULE024 RULE02S RULE026 RULE027 RULE028 

RULE029 RULE030 RULE031 RULE032 RULE046 
RULE047 RULE048) 

USED-BY: (RULE049 RULEOSO RULEOSl) 
UPDATED-BY: (RULEOOl RULE002 RULE003) 

FRL [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (footrest lower height in centimeters) 
EXPECT: POSNUMB 
DICTIONARY: INTERNAL 
UPDATED-BY: (RULE024 RULE02S RULE026 RULE027 RULE028 

RULE029 RULE031 RULE032 RULE046 RULE047 
RULE048) 

FRU [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (footrest upper height in centimeters) 
EXPECT: POSNUMB 
DICTIONARY: INTERNAL 
UPDATED-BY: (RULE024 RULE02S RULE026 

RULE029 RULE030 RULE031 
RULE047 RULE048) 

RULE027 
RULE032 

RULE028 
RULE046 

HEADER [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (*) 
EXPECT: ANY 
UPDATED-BY: (RULE0S2 RULE053 RULE054 RULEOSS) 

LARGE-AREA [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (the requirement to cover a large work area) 
PROMPT: (Does the short-cycle performance of this task 

require the operator to cover a large work area?) 
REPROMPT: (Is the operator required to get up and move 

about a large work area frequently to stock 
parts, check equipment, etc. as part of his 
duties at the workplace?) 



86 

ASKFIRST-: T 
USED-BY: (RULE004 RULEOOS RULE006 RULE007 RULE008 RULE009 

RULEOIO RULEOll RULE012 RULE013 RULE014 RULEOIS 
RULE016 RULE017 RULE018 RULE019) 

MALE-DATA-ERROR [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (there is a probable error in the male 
anthropometric data) 

USED-BY: (RULE0S2 RULEOSS) 
UPDATED-BY: (RULE049) 

MHR [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (maximum horizontal reach in centimeters) 
EXPECT: POSNUMB 
DICTIONARY: INTERNAL 
UPDATED-BY: (RULE024 RULE025 RULE026 RULE027 RULE028 

RULE029 RULE030 RULE031 RULE032 RULE046 
RULE047 RULE048) 

MIXED-DATA-ERROR [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (there is a probable error in the mixed 
anthropometric data) 

USED-BY: (RULE054 RULEOSS) 
UPDATED-BY: (RULEOSl) 

MU-HAT [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (the estimated thigh width mean) 
EXPECT: POSNUMB 
USED-BY: (RULE035 RULE036 RULE037 RULE038 RULE039 RULE040 

RULE041) 
UPDATED-BY: (RULE034) 

MVR [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (maximum vertical reach in centimeters) 
EXPECT: POSNUMB 
DICTIONARY: INTERNAL 
UPDATED-BY: (RULE024 RULE025 RULE026 RULE027 RULE028 

RULE029 RULE030 RULE031 RULE032 RULE046 
RULE047 RULE048) 



87 

NHR [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (normal horizontal reach in centimeters) 
EXPECT: POSNUMB 
DICTIONARY: INTERNAL 
UPDATED-BY: (RULE024 RULE025 RULE026 RULE027 RULE028 

RULE029 RULE030 RULE031 RULE032 RULE046 
RULE047 RULE048) 

NVR [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (normal vertical reach in centimeters) 
EXPECT: POSNUMB 
DICTIONARY: INTERNAL 
UPDATED-BY: (RULE024 RULE02S RULE026 RULE027 RULE028 

RULE029 RULE030 RULE031 RULE032 RULE046 
RULE047 RULE048) 

POPLITEALS [DIMENSIONAL-DESIGN-PARMS] 

TRANS: 
PROMPT: 

REPROMPT 

(the Sth percentile popliteal height) 
(For your specific population, enter the SMALLEST 
Sitting Popliteal Height in centimeters.) 

: (The Sitting Popliteal Height - is the vertical 
distance from the floor to the underside of the 
thigh immediately behind the knee.) 

POSNUMB EXPECT: 
CONTAINED-IN: (RULE024 RULE026 RULE027 RULE029 RULE030 

RULE032 RULE034 RULE046 RULE048) 
USED-BY: (RULE049 RULEOSO RULEOSl) 
UPDATED-BY: (RULEOOl RULE002 RULE003) 

P0PLITEAL95 [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (the 95 th percentile popliteal height) 
PROMPT: (For your specific population, enter the LARGEST 

Sitting Popliteal Height in centimeters.) 
REPROMPT: (The Sitting Popliteal Height - is the vertical 

distance from the floor to the underside of the 
thigh immediately behind the knee.) 

EXPECT: POSNUMB 
CONTAINED-IN: (RULE024 RULE027 RULE030 RULE034 RULE046 

RULE048) 
USED-BY: (RULE049 RULEOSO RULEOSl) 
UPDATED-BY: (RULEOOl RULE002 RULE003) 



88 

POPSPEC [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (the sex of your specific population) 
PROMPT: (In regard to the anthropometric measurements that 

you entered, what is the sex of your specific 
work force population?) 

ASKFIRST: T 
EXPECT: (MALE FEMALE MIXED) 
USED-BY: (RULE0S2 RULE0S3 RULE0S4) 

POPULATION [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (the workforce population) 
PROMPT: (Describe the population of people who will use 

this workplace as: :line GENERAL MALE POPULATION 
- Sth to 95 th male percentiles :line GENERAL 
FEMALE POPULATION - Sth to 95 th female 
percentiles :line GENERAL MIXED POPULATION - Sth 
percentile female to 95 th percentile male :line 
YOUR SPECIFIC POPULATION - anthropometric 
measurements entered by you.) 

ASKFIRST: T 
EXPECT: (MALE FEMALE MIXED SPECIFIC) 
USED-BY: (RULEOOl RULE002 RULE003 RULE043 RULE044 RULE04S 

RULE0S2 RULE053 RULE054 RULEOSS) 

POPULATION-REDUCTION [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (the approximate reduction to the population that 
is to be accommodated at this workplace, due to a 
thigh clearance problem) 

EXPECT: (NO-REDUCTION <S% 5-10% ilO-20%^ i20-30%1[ i30-40%«i[ 
i40-50%11 >50%) 

DICTIONARY: INTERNAL 
UPDATED-BY: (RULE035 RULE036 RULE037 RULE038 RULE039 

RULE040 RULE041 RULE042 RULE033) 

SHOULDERS [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (the Sth percentile standing shoulder height) 
PROMPT: (For your specific population, enter the SMALLEST 

Standing Shoulder Height in centimeters.) 
REPROMPT: (The Standing Shoulder Height - is the vertical 

distance from the floor to the upper-most point 
on the lateral edge of the shoulder with the 
operator standing erect.) 

EXPECT: POSNUMB 
CONTAINED-IN: (RULE025 RULE026 RULE028 RULE031 RULE032 

RULE047) 



89 

USED-BY: (RULE049 RULEOSO RULEOSl) 
UPDATED-BY: (RULEOOl RULE002 RULE003) 

SH0ULDER9S [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (the 95 th percentile standing shoulder height) 
PROMPT: (For your specific population, enter the LARGEST 

Standing Shoulder Height in centimeters.) 
REPROMPT: (The Standing Shoulder Height - is the vertical 

distance from the floor to the upper-most point 
on the lateral edge of the shoulder with the 
operator standing erect.) 

EXPECT: POSNUMB 
USED-BY: (RULE049 RULEOSO RULEOSl) 
UPDATED-BY: (RULEOOl RULE002 RULE003) 

SIGMA-HAT [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (the estimated thigh width standard deviation) 
EXPECT: POSNUMB 
USED-BY: (RULE03S RULE036 RULE037 RULE038 RULE039 RULE040 

RULE041) 
UPDATED-BY: (RULE034) 

SITTING-ELBOWS [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (the Sth percentile sitting elbow height) 
PROMPT: (For your specific population, enter the SMALLEST 

Sitting Elbow Height in centimeters.) 
REPROMPT: (The Sitting Elbow Height - is the vertical 

distance from the sitting surface to the bottom 
of the elbow.) 

EXPECT: POSNUMB 
CONTAINED-IN: (RULE024 RULE026 RULE027 RULE029 RULE030 

RULE032 RULE046 RULE048) 
USED-BY: (RULE049 RULEOSO RULEOSl) 
UPDATED-BY: (RULEOOl RULE002 RULE003) 

SITTING-ELB0W95 [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (the 95 th percentile sitting elbow height) 
PROMPT: (For your specific population, enter the LARGEST 

Sitting Elbow Height in centimeters.) 
REPROMPT: (The Sitting Elbow Height - is the vertical 

distance from the sitting surface to the bottom 
of the elbow.) 

EXPECT: POSNUMB 
CONTAINED-IN: (RULE024 RULE026 RULE027 RULE029 RULE030 



90 

RULE032 RULE046 RULE048) 
USED-BY: (RULE049 RULEOSO RULEOSl) 
UPDATED-BY: (RULEOOl RULE002 RULE003) 

SITTING-SHOULDERS [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (the Sth percentile sitting shoulder height) 
PROMPT: (For your specific population, enter the SMALLEST 

Sitting Shoulder Height in centimeters.) 
REPROMPT: (The Sitting Shoulder Height - is the vertical 

distance from the sitting surface to the 
upper-most point on the lateral edge of the 
shoulder with the operator sitting erect.) 

EXPECT: POSNUMB 
CONTAINED-IN: (RULE024 RULE027 RULE029 RULE030 RULE046 

RULE048) 
USED-BY: (RULE049 RULEOSO RULEOSl) 
UPDATED-BY: (RULEOOl RULE002 RULE003) 

STATION-TYPE [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (the workstation type) 
EXPECT: (SIT STAND SIT/STAND) 
DICTIONARY: INTERNAL 
USED-BY: (RULE024 RULE025 RULE026 RULE027 RULE028 RULE029 

RULE030 RULE031 RULE032 RULE046 RULE047 RULE048 
RULE042 RULE0S7) 

UPDATED-BY: (RULE004 RULEOOS RULE006 RULE007 RULEOOS 
RULE009 RULEOIO RULEOll RULE012 RULE013 
RULE014 RULEOIS RULE016 RULE017 RULEOIS 
RULE019) 

TABLE-THICKNESS [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (the table thickness) 
PROMPT: (Enter the work surface table thickness in 

centimeters.) 
REPROMPT: (The thickness of the work surface table is the 

vertical depth dimension of the table that is 
to be used as the work surface.) 

ASKFIRST: T 
EXPECT: POSNUMB 
CONTAINED-IN: (RULE034) 
USED-BY: (RULE0S6 RULE057) 

THIGH-CRITICAL [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (the critical thigh-clearance height) 



91 

EXPECT: POSNUMB 
USED-BY: (RULE035 RULE036 RULE037 RULE038 RULE039 RULE040 

RULE041) 
UPDATED-BY: (RULE034) 

THIGH-CUT [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (the thigh clearance difference) 
EXPECT: POSNUMB 
UPDATED-BY: (RULE034) 

THIGH-POPS [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (the Sth percentile thigh-width / popliteal 
combination) 

EXPECT: POSNUMB 
UPDATED-BY: (RULE034) 

THIGH-P0P9S [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (the 95 th percentile thigh width / popliteal 
combination) 

EXPECT: POSNUMB 
UPDATED-BY: (RULE034) 

THIGH-TEST [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (there is a thigh clearance problem) 
USED-BY: (RULE034 RULE033) 
UPDATED-BY: (RULE0S6 RULE0S7) 

THIGH-WIDTHS [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (the Sth percentile thigh width) 
PROMPT: (For your specific population, enter the SMALLEST 

thigh width measurement in centimeters.) 
REPROMPT: (The Thigh Width - is the vertical distance from 

the sitting surface to the top of the thigh at 
its intersection with the abdomen.) 

EXPECT: POSNUMB 
CONTAINED-IN: (RULE034) 
USED-BY: (RULE049 RULEOSO RULEOSl) 
UPDATED-BY: (RULE043 RULE044 RULE045) 



92 

THIGH-WIDTH95 [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (the 95 th percentile thigh width) 
PROMPT: (For your specific population, enter the LARGEST 

thigh width measurement in centimeters.) 
REPROMPT: (The Thigh Width - is the vertical distance from 

the sitting surface to the top of the thigh at 
its intersection with the abdomen.) 

EXPECT: POSNUMB 
CONTAINED-IN: (RULE034) 
USED-BY: (RULE049 RULEOSO RULEOSl RULE0S6 RULE057) 
UPDATED-BY: (RULE043 RULE044 RULE04S) 

WORK-CLASS [DIMENSIONAL-DESIGN-PARMS] 

TRANS: 
PROMPT: 

REPROMPT 

ASKFIRST 
EXPECT: 

USED-BY: 

(the classification of the task) 
(How would you classify the tas 
performed at this workplace?) 

: (Example tasks categorized by 
Classification: :line :line 
Inspection, Fine Assembly, S 
:line LIGHT WORK - Manual As 
of Machine, Objects < 4.5 kg 
- Packing, Wrapping, Objects 
VDT/KEYBOARD OPERATION - Dat 
Secretarial, Programming, et 

: T 
(PRECISION-WORK LIGHT-WORK HEAV 
VDT/KEYBOARD-OPERATION) 
(RULE004 RULEOOS RULE006 RULEO 
RULEOIO RULEOll RULE012 RULEO 
RULE016 RULE017 RULEOIS RULEO 
RULE022 RULE023) 

k that is to be 

Work 
PRECISION WORK -
oldering, etc. 
sembly, Load/Unload 
. :line HEAVY WORK 
> 4 . 5 kg. :line 

a Processing, 

c.) 

Y-WORK 

07 RULEOOS 
13 RULE014 
19 RULE020 

RULE009 
RULEOIS 
RULE021 

WSL [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (work surface lower ht. in centimeters) 
EXPECT: POSNUMB 
DICTIONARY: INTERNAL 
CONTAINED-IN: NIL 
USED-BY: NIL 
UPDATED-BY: (RULE024 RULE02S RULE026 RULE027 RULE028 

RULE029 RULE030 RULE031 RULE032 RULE046 
RULE047 RULE048) 

WSU [DIMENSIONAL-DESIGN-PARMS] 

TRANS: (work surface upper ht. in centimeters) 
EXPECT: POSNUMB 



93 

DICTIONARY: INTERNAL 
CONTAINED-IN: (RULE034) 
USED-BY: (RULE056 RULE0S7) 
UPDATED-BY: (RULE024 RULE025 RULE026 RULE027 RULE028 

RULE029 RULE030 RULE031 RULE032 RULE046 
RULE047 RULE048) 

Domain Variables 

$$TITLE [DOMAIN.VARIABLES] 

^^^^^- (" . ,, THE ERGONOMIST" :line 
•li^^ Workplace Design Expert 
System" :line :line :line :line :line " 
copyright Thomas B. DeGreve 1985") 

System parameters 

DOMAIN [SYSVARS] 

Value: "THE ERGONOMIST" 

TREEROOT [SYSVARS] 

Value: DIMENSIONAL-DESIGN 

System parameters 

ANTHRO-RULES [RULEGROUPS] 

SVAL: (ANTHROPOMETRIC DATA) 
CONTEXT: (DIMENSIONAL-DESIGN) 
Value: (RULEOOl RULE002 RULE003 RULE043 RULE044 RULE045) 

DATA-CHECKING-RULES [RULEGROUPS] 

CONTEXT: (DIMENSIONAL-DESIGN) 
SVAL: (DATA CHECKING) 
Value: (RULE049 RULEOSO RULEOSl RULE052 RULE053 RULE054 

RULEOSS) 

DIMENSIONING-RULES [RULESGROUPS] 

CONTEXT: (DIMENSIONAL-DESIGN) 
SVAL: (WORKPLACE DIMENSIONS) 
Value: (RULE020 RULE021 RULE022 RULE023 RULE024 RULE025 



94 

RULE026 RULE027 RULE028 RULE029 RULE030 RULE031 
RULE032 RULE046 RULE047 RULE048) 

STATION-TYPE-RULES [RULEGROUPS] 

CONTEXT: (DIMENSIONAL-DESIGN) 
SVAL: (WORKSTATION TYPE) 
Value: (RULE004 RULEOOS RULE006 RULE007 RULEOOS RULE009 

RULEOIO RULEOll RULE012 RULE013 RULE014 RULEOIS 
RULE016 RULE017 RULEOIS RULE019) 

THIGH-CLEARANCE-RULES [RULEGROUPS] 

CONTEXT: (DIMENSIONAL-DESIGN) 
SVAL: (THIGH CLEARANCE) 
Value: (RULE033 RULE034 RULE03S RULE036 RULE037 RULE038 

RULE039 RULE040 RULE041 RULE042 RULE0S6 RULE057) 

System parameters 

CONTEXTTYPES [PARMGROUPS] 

Value: (DIMENSIONAL-DESIGN) 

DIMENSIONAL-DESIGN-PARMS [PARMGROUPS] 

Value: (WORK-CLASS POPULATION STATION-TYPE WSL CHL CHU FRL 
FRU NVR MVR NHR MHR WSU ARMS FOREARMS B0DY-DEPTH9S 
SITTING-SHOULDERS SITTING-ELB0W95 POPLITEALS 
P0PLITEAL9S SHOULDERS SH0ULDER9S ELBOWS ELB0W9S 
SITTING-ELBOWS C2 CI DESIGN-TYPE FOOT-CONTROLS 
LARGE-AREA TABLE-THICKNESS THIGH-WIDTH9S 
POPULATION-REDUCTION THIGH-CUT MU-HAT SIGMA-HAT 
THIGH-WIDTHS THIGH-P0P95 THIGH-CRITICAL THIGH-POPS 
MALE-DATA-ERROR FEMALE-DATA-ERROR MIXED-DATA-ERROR 
POPSPEC HEADER THIGH-TEST) 



APPENDIX C 

SAMPLE CONSULTATIONS 

• Knowleage base .*: iHt tRGuNunrsT ^ 

CURRENT OBJECTIVE: 

THE ERGONOMIST is a knowledge-based workplace design program 
developed to provide sound erqonomic advice to the' 
industrial workstation designer. 

The current objective is to: 
(1) Decide what type of workstation to design and 
(2) Provide critical dimensions required to construct the 
workstation and 
(3) Determine it* there is a reduction in the accommodated 
population due to physical constraints. 

... end — press RETURN 

Up Down CF Unknwn Done — Why How Help Undo Stoo 
, Fl F2 F3 F4 F5 F6 F7 F3 F9 FIO ESC , 

How would you classify the task that is to be performed at 
this workplace? 
Select one of the following: (Press T 3 ' for selection help.) 

> PRECISION-WORK 
LIGHT-WORK 
HEAVY-WORK 
VDT/KEYBOARD-OPERATION 

Up Down CF Unknwn Done — Why How Help Undo Stop 
Fl F2 F3 F4 F5 Fb Ff F8 F9 FIO ESC 

95 



Describe the population of people who will use this 
workplace as: 
GENERAL HALE POPULATION - Sth to 95 th male percentiles 
GENERAL FEMALE POPULATION - Sth to 95 th female percentiles 
GENERAL MIXED POPULATION - Sth percentile female to 95 th 
percentile male 
YOUR SPECIFIC POPULATION - anthropometric measurements 
entered by you. 
Select one of the following: (Press T 3 ' for selection help.) 

^ MALE 
>FEMALE 

MIXED 
SPECIFIC 

Uc 

I 

p uown LP Unknwn Done — Why How Help Undo Stop 
. Fl F2 F3 F4 F5 F6 FT F8 F9 FIO ESC 

knowledge 

For this task, is the operator required to use any foot 
controls? 
Select one of the following: (Press ^F3' for selection help.) 

>YES 
NO 

Uc Stot p Down CF Unknwn Done — Why How . Help Undo stop 
. Fl F2 F3 F4 F5 F6 FT F8 F9 FIO ESC 

Does the short-cycle performance of this task require the 
operator to cover a large work area? 
Select one of the following: (Press T S " for selection help.) 

YES 
> N 0 

UG Down 
F2 

CF Unknwn 
F3 F4 

Done 
F5 

— 

F6 fF How F8 Help Undo F9 FIO Stop ESC 
^^^A^^^^^^^MW^/v^M^^A^^MVM^MM^^MM^n^^MWinA'/jv^jvwd 

96 



97 

What type of design do you wish to construct? 

FULLY ADJUSTABLE - with adjustable worksurface, chair, and 
footrest 
FlXg WpRK^SURFACE - with adjustable chair and footrest 
PIJED (>WIR - with adjustable worksurface and footrest 
STATIONARY - with no adjustable features 
Select one of the following: (Press T 3 ' for selection help.) 

FULLY-ADJUSTABLE 
^ FIXED-WORK-SURFACE 
> FIXED-CHAIR 

STATIONARY 

Up Down CF Unknwn Done — Why How Help Undo Stop 
, Fl F2 F3 F4 F5 F6 F T F8 F9 FIO ESC 

MMFLACE-1 is as follows: The following conclusions were made 
by THE ERi^NOHIST concerning the current workplace design. 

The workstation type is as follows: SIT 

Work surface upper ht. in centimeters is as follows: 83.0 

Work surface lower ht. in centimeters is as follows: 74.4 

Chair upper height in centimeters is as follows: 43.6 

Chair lower height in centimeters is as follows: NOI€ 

Footrest upper height in centimeters is as follows: 6.3 

I was unable to make any conclusions regarding footrest 
lower height in centimeters. 

Normal vertical reach in centimeters is as follows: 94.9 

Maximum vertical reach in centimeters is as follows: 156.0 

Normal horizontal reach in centimeters is as follows: 18.7 

Maximum horizontal reach in centimeters is as follows: 46.4 

The approximate reduction to the population that is to be 
accommodated at this workplace, due to a thigh clearance 
problem is as follows: NO-REDUCTION 



CURRENT OBJECTIVE: 

TIC ERGONOMIST is a knowledge-based workplace design program 
developed to provide sound ergonomic advice to the 
industrial workstation designer. 

The current objective is to: 
1 Decide what type of workstation to design and 
(2) Provide critical dimensions required to construct the 
workstation and 
(3) Determine if there is a reduction in the accommodated 
population due to physical constraints. 

... end — press RETURN 

98 

Up Down CF Unknwn Done — Why How Help Undo Stop 
, Fl F2 F3 F4 F5 F6 FT F8 F9 FIO ESC 

How would you classify^ the task that is to be performed at 
this workplace? 
Select one of the following: (Press ''F3' for selection help.) 

PRECISIOfHJORK 
LIGHT-WORK 

>HEAVY-WORK 
VDT/KEYB0AR1>-0PERATI0N 

Up Down CF Unknwn Done — . W h y How 
, Fl F2 F3 F4 F5 F6 FT F8 ., . ., 

Help Undo 
F9 FIO 

r.v//.\wi 

Describe the population of people who will use this 
workplace as: __ 
GENERAL MALE POPULATION - Sth to 95 th male percentiles 
GENERAL FEMALE POPULATION - Sth to 95 th female percentiles 
GENERAL MIXED POPULATION - Sth percentile female to 95 th 
percentile male 
YOUR SPECIFIC POPULATION - anthropometric measurements 
entered by you. 
Select one of the following: (Press T 3 ' for selection help.) 

>MALE 
FEMALE 
MIXED 
SPECIFIC 

Up Down CF Unknwn Done — wny 
Fl F2 F3 F4 F5 F6 FT jFS F9 ' FIO ESC 

Why How ^too Help Undo 
^ n r^ rj r«t rg FQ r/ jro F9 FIO coî  ^ 



99 

For this task, is the operator required to use any foot 
controls? 
Select one of the following: (Press ^F3' for selection help.) 

^ YES 
> N 0 

Up Down CF Unknwn Done — Why How Help Undo Stop 
Fl F2 F3 F4 P5 F6 FT F8 F9 FIO ESC 

• Know ledge Base :: THt tRGuNUnrbT ^ 

Does the short-cycle performance of this task require the 
operator to cover a large work area? 
Select^one of the following: (Press ^F3' for selection help.) 

>YES * 
NO 

Up Down CF Unknwn Done — Why How Help Undo Stop 
, Fl F2 F3 F4 F5 F6 FT F8 F9 FIO ESC 

What type of design do you wish to construct? 

FULLY ADJUSTABLE - with adjustable worksurface, chair, and 
footrest . , . . 
FIXED WORK SURFACE - with adjustable chair and footrest 
FIXED CHAIR - with adjustable worksurface and footrest 
STATIONARY - with no adjustable features , , .̂ ^ , ^ 
Select one of the following: (Press T 3 ' for selection help-.) 

FULLY-ADJUSTABLE 
FIXED-WORK-SURFACE 
FIXED-CHAIR 

>STATIONARY 

Up Down CF Unknwn Done — Why How Help Undo Sto 
, Fl F2 F3 F4 F5 F6 F7' F8 F9 FIO £SL 



100 

*"«SM5a55'ifi5?«'fln WAW/WWJWA'A'JWf 
Enter the work surface table thickness in centimeters. 
Enter a positive number. 

10 

Down CF Unknwn Up uown u- unKnwn uone — wny 
Fl F2 F3 F4 F5 F6 F T 

Done Wh\ How Help Undo 
F9 FIO 

Sto M 
W//^JVWJV/J 

WORKPLACE-1 is as follows: The following conclusions were made 
by Tl£ ERGONOMIST concerning the current workplace design. 

The workstation type is as follows: STAND 

Work surface upper ht. in centimeters is as follows:. 107.4 

Work surface lower ht. in centimeters is as follows: NONE 

Chair upper height in centimeters is as follows: NONE 

Chair lower height in centimeters is as follows: NOI€ 

Footrest upper height in centimeters is as follows: 7.4 

Footrest lower height in centimeters is as follows: NOfE 

Normal vertical reach in centimeters is as follows: 150.1 

Maximum vertical reach in centimeters is as follows: 213.7 

Normal horizontal reach in centimeters is as follows: 20.1 

Maximum horizontal reach in centimeters is as follows: 51.3 

The approximate reduction to the population that is to be 
accommodated at this workplace, due to a thigh clearance 
problem is as follows: NO-REDUCTION 



y iCnowieoge Base '•'• i r t twOuNUrfriT ^ 

J CURRENT OBJECTIVE: 

J THE ERGONOMIST is a knowledqe-based workplace design proaram 
, developed to provide sound ergonomic advice to the' 
^ industrial workstation designer. 
y 

J The current objective is to: 
. (1) Decide what type of workstation to design and 
. (2) Provide critical dimensions required to construct the 
r workstation and 
. (3) Determine if there is a reduction in the accommodated 
^ population due to physical constraints. press RETURN 

h 
h 
h 

« 

U f Down F2 CF Unknwn F3 F4 Done F5 F6 
Whv How 

F8 
Help Undo 
F9 FIO 

Stoo 
ESC 

.•WAV^MV.VAVA"A"/WA'A"/^.VAV/AV.V/^/MV/AV.V.»AV.V.V.\VAV.*JVAV.V.VJ 

^.VV'jTANVV/AW/AV.-M^AVAViffVi'JAVAVAS'AV^^^^ 
r Ĵiowiedqe Base :: THt tKbiiNunrsi ? 
J How would you classify the task that is to be performed at 
• this workplace? 
• S e l e c t one of the followinq: (Press 'FS' for selection help.) 
•• PRECISION-WORK 

LIGHT-WORK 
HEAVY-WORK 
VDT/KE'fBOARD-OPERATION 

Stop 
ESC 

J Up Down CF Unknwn Done — Why How >He!0 Undo 
J Fl F2 F3 F4 FS F6 F7^ F3 F9 -10 

•V/^WAV//^M^WAV/^^*VAVAVA»AV^/A»^AVAVAWiVAVA»AVA»AV.VA«AVAVA'JIl 

>AVAVAViV;'A»ANVAWA"iCi%Y,!A%%V^^AlAVAVA»A'.VAVA%V."^.V/AVA%VAVA»^ 

7 Knowledge case :: THE tRGUNijnibi 

J Example tasks categorized by Work Classification: 

?• PRECISION WORK - Inspection, Fine Assembly,'Soldering, jtc. 
r LIGHT WORK - Manual Assembly, Load/Unload of Hachinel 
• Objects C 4.5 kq. 
* HEHVY WORK - Packing. Wrapping. Objects > ̂ .5 kg. 
J VDT/KEYBOARD OPERATION - Data'Processing, Secretarial, 
r Programming, etc. 
h 
k 
h 
h 
k 
h 
k 
k 
k 
k 
k 
t 

elect one of the following: (Press T3'' for selection help.) 
PRECISION-WORK 
LIGHT-WORK 

^ HEAVY-WORK 
> VDT/KE/B0ARDH3PERATI0N 

Up 
Fl 

Down 
C-7 

CF Unknwn 
F3 F4 

L'one 
F5 F6 

Whv 
F7' 

How 
F8 

He ID 'Jndo 
F9 FIO 

zzo:; 

.•AV/AWiTAW.VAV.VAWAV.VAV.V.'AVJ'̂ .VAV̂ AV.' 
I 

•J 

101 



J'A'AVAV.yA-ANV.V.yCAVA&W.ViW/AV.VAW^^^^ 
,. N.r.cijji3gge case : : IHE tNijuNtjMiST 7 
Describe the population of people who will use this 
aorKpiace as: 
GENERAL ?1ALE ?i3PULATI0N - 5th to 95 th male percentiles 
IJENERAL FEMALE POPULATION - Sth to 95 th female percentile^ 
GENERAL MIXED POPULATION - Sth percentile female to 95 th 
percentile male ' 
YOUR SPECIFIC POPULATION - anthropometric measurements 
entered by you. 
Select one of the following: (Press 'FS' for selection help.) 

MALE 
^FEMALE 
> MIXED 

102 

SPECIFIC 

Done 
F5 

— Why How 
F6 F7^ F8 

Help Undo 
F9 FIO 

StOD 
ESC 

Up Down CF Unknwn 
Fl F2 r3 F4 

VAVAVA%VA"AV/AVJ'J'^AVA'/A'.VAVAr/J»d'/^/^A'AV^AVAVA\VAWJV/AV,V.V^^^ 

• m 

For this task, is the operator required to use any foot 
controls? 
Select one of the following: (Press ''FS' for selection help.) 

YES 
> N 0 

Done 
F5 

— -

F6 
Why How 

F8 
Help undo 
F9 FIO 

Stoo 
ESC 

Up Down CF Unknwn 
Fl F2 F3 F4 

AVAVA»J"^iPJ'//^.»A'A»tV/A'AVAV/AViV^AVAVAW/AV.'yVAV^/.^A»//A».V^/i^AVA*J 

Does the short-cycle performance of this task require the 
operator to cover a large work area? 
S-̂ iect one of the following: (Press "FS' for selection help.) 

> N 0 

JC'ijjn ijnKnwn 
-4 

Done 
^5 

— Why How 
F6 n' Fo 

AVA'AVAVAV.V 

nelo 'jrico i t CO 

•i 



103 
Ĵ VA%%VytVA'AVAVAWA»<A%V»ViVW»VAVA%«̂ ^̂ ^̂ ^ 
• Knowledgebase :: l^tRedfltjKiSi * 

What type of design do you wish to construct? 

PJLLY AD.JUSTABLE - with adjustable worksurface, chair, and 
footrest 
FIXED WORK SURFACE - with adjustable chair and footrest 
FIXED CHAIR - with adjustable worksurface and footrest 
STATIONARY - with no adjustable features 
Select^one of the following: (Press "FS" for selection help.) 

> FULLY-ADJIJSTABLE 
FIXED-WORK-SURFACE 
FIXED-CHAIR 
STATIONARY 

r^ Down F2 CF Unknwn F3 F4 Done F5 F6 Why F7 How F8 Help Undo F9 FIO Stop ESC 
.W.W/AV/A%ViVJ'/^/AV^^AWJ'A»/A"JVJ»ir/A"AW/^AV^/J"A%VAVAVAVAV.VAV.'J 

WORKPLACE-1 is as follows: The following conclusions were made 
by THE ERGONOMIST concerning the currenfworkplace design. 

The workstation type is as follows: SIT 

Work surface upper ht. in centimeters is as follows: 76.1 

Work surface lower ht. in centimeters is as follows: 56.1 

Chair upper height in centimeters is as follows: 48.7 

Chair lower height in centimeters is as follows: 37.3 

Footrest upper height in centimeters is as follows: 11.4 

Footrest lower height in centimeters is as follows: 0 

Normal vertical reach in centimeters is as follows: 88.6 

Maximum vertical reach in centimeters is as follows: 149.7 

Normal horizontal reach in centimeters is as follows: 15.5 

Maximum horizontal reach in centimeters is u follows: 43.2 

The 
acco 
pr 

e approximate reduction to the population that is to be 
commodated at this workplace, due to a thiqh clearance 
obi em is as follows: NO-REDUCTION 



^ ..nojjiijqe case i : 'rE Eh;buNOni';i . 

CURRENT DSJEC'IVE: 

'rtE ERGONOMIST is a knowledge-based 'oiorkplace design program 
developed to provide sound ergonomic idvici to '"''»' 
industrial workstation designer. 

The current objective is to: 
•U S®*^ide what.type of workstation to design and 

ui Provide cntitai dimensions required to'construct the 
•^lorkstation and 
13) Determine if there is a reduction in the accommodated 
population due to physical constraints. 

... end — press RETIJRN 

UP Down CF Unknwn 
Fl F2 F3 F4 

Done 
FS 

— Why How 
F6 F7' 

He ID Undo ^m 
AVt"A'.VAV.VbV/A"AV^/^/AVA%V.VA«A'A%VA%VA%VAVAVASWAVAVAV.V.VAVJ 

J'.'AVVAVNy.VAV.V.WiVA'iAiV.W.W^^ 
• •̂ •nowiedge Base :: iHE arnwIoHrSi ^ 

How would you classify the task that is to be perforTied at 
this workplace? 
Select one of the followinq: (Press 'FS'' for seiecTion help.) 

^ PRECISION-WORK 
> LIGHT-WORK 

HEAVY-WORK 
VDT/KE'/BOARD-OFERATION 

Up Down CF Unknwn Done 
Fl F2 F3 F4 F5 

— Whv How 
F6 F7^ F3 

^iiD Undo • - , ) • • • " 
":-r 

I. 
/A"AVi'AV/AVAV«»AVJ'J'AV///AVAViVJ»AVAS%VAVAVAVA'A"AVAVAV.V.'.VAVJ 

^VAVA^VA•A^^VAVA'A•^iVA>^y^VAVAVAVAVAVAV.^^^VAVA^^v.^^^'^^^^^^^ 
K.nvJjjiecce case '•' THE tKbijNtjniji 

Describe the population of pec-p:̂  -jho :ni:! use this 
wor-itD'ace as: 
'3ENEF:AL flALE POPULATION - Sth t : 95 th inai^ p-i-Cirtiies 
GE'MERAL FEMALE 'POPULATION - Sth t:i 'ij th femaie osf^centi.es 
GENERAL HIXED POPULATION - oth oercantiie female to ^5 t" 
percentile male 
YOUR SPECIFIC POPULATION - anthropometric measurements 
entered py vou. 
Select one of the fclloniina: iPf̂ ess "'Fl' for selection h e b . • 

<̂ALE 

> (IIXED 
cpcr TPTP 

UD Do'a/ri CF unKnwn Done - ~ ^hy 
Fl *2 Fl r^ F5 Fk p " 

' i i - 11' 

I 

104 

1 I I f t 1 1 1 t i t I I f • i i i t t i i i i i i a i i i i J i i i i i i i i i i i i i i i t i t i i i t t i i i . i i i i i i i i t t t i i i i i i i a l 



|AVAN%SVA%%yAVA%Vyfy,!,yy/,W/AVAVAVAV.VAV,V,%VAV.V.SV.V.VAV.V/AV^ 

• ^r!Owledge base :: \Kt iKi-UNUMi;;:i ^ 
J For t n i s t i S K , is trie cp<iraror r e q u i r e d to use any foot 

7 -lontrois*' 
"Select one of the followinq: (Press 'F3' for seiectior'. r,ai:'.f 
I NO' 
fa 
k 
k 
k 
k 
k 
k 
k 
k 
k 
k 
k 
k 
Y Up Down CF Unknwn Done — > Why How 
J Fl F2 F3 F4 FS F6 F7" F3 
.•A'.VAVA"AVASSVAVAVAV.VASVA%VAVAViV.VAVAViVAV.VAV.VAV,VtV,VAVJ 

>tVAViV/iyi^/J^AVAVA'A'.Vi%V«iVjV«VAVAVAVAVAV.VAVAV.VAV.V.'^ 
J KT.cwie'̂ geBase : : I h ^ crtiJj^jdnriT z. 

J For this task, is the operator required to use anv foot 
r controls? 
f Why this question is needed: 

105 

F? FiO 
•'129 

^/Jhether the requirement to *. 
* the w o r k s t a t i o n type 

use f o o t controls i s needed to determine 

; RULEOOS 
J If 1) the classification of the task is LIGHT-WORK, and 
* 2) the requirement to use foot controls, and 
* 2) the requirement to cover a larqe aorK area. 
* Then it is definite (1007.) that the workstation tvpe is 
J SIT/STAND. 

I ... end - press RETURN 
k 
k 
k 

»• lip 
" F l 

Down CF Unknwn 
F2 F4 

Done - - - Why How Heip J^J 
fO F7 l-V 

\vAVA'iPAVAV.VA'^^^^AVAVAVAV.VAVAV.VAyAVAViVAV.y,y.y.V.V.yAVA%^^^ 

>AVAVAVyA'C^AV.V.-AW4i,>ytW^ViW'*"W^^^ 
Knowî qqe aase : : irE tnijijTTiJMibi ^ 

** For this task, is the operator required to use my foot 
^ "on^Tois? 
^ Sei^'-t one of the followinq: (Press "FS-' tor selection nel?. 
J >N0 
k 
k 
k 
k 
k 
k 
k 
k 
k 
k 
k 
k 
k 
^ l!p D''iyri i"F 'Til-nujn Done — '-'Jhv roiij 
fc pi ' r - r-.-, F4 

Done — ''Jhv 
F5 F6 F7' 

.'r j . : 
- : u 

. . . • • • • • • • . . . • • . . • • • • • • • • i i i i i i i i i i i * i i ' * * * * ' * * * * * * * ' * ' * ' * i * * i i i * * * < i * * i i i i i i 1 g V i i i V i V i i 1 1 1 1 i / t ^ i y A A i n I 1 1 1 I • i i i t i i i i i n i i t i i i i i i i i i t i i i i t i t t i i i i i i i i . i j i twd 



^AVAVAVpVi*AVAVA'A'A'AiyAViV»NViVAy.VAVAV.VAV.VAy.'.yi'.V.V.V.'.VAy.J 
• r.riowieqge sase '•'• :rtE-.-.biJwjniST ^ 
^ Does the short-cycle jerforn^ance of this task requiri: t'l^ 
. operator to cover a large 'jjork area? 
. S e l e c t one of tr.e foliowinq: (Press "F3' for S'jiection n e l p . ' 

J >M0 
k 
k 
k 
k 
k 
k 
k 
k 
k 
k 
k 
k 
k 

J Up Down CF Unknwn Done — Why How Help Undi 
J Fl F2 F3 -4 FS F6 F7' ^3 F9 FIO 
ASV.»AViy.V//A%SSViV.VAVAV^AVAVAV.V«V/AVAVAVAVAVAViVAVAVAyAVi 

J'AVAVAV/Asiv.VAVA*AyAyjCf!«iViyAv,v.yiViy,y.v.v.vAv.y,VAV.yAv.«.yAv| 

J knowledge ease :: i H r ^ r a j ? u : ? i ^ 

^ What type of design do you wish t: constr'jct'' 

J RJLL/ AD'JUSTABLE - with adjustable work:sur-ace, chai:-. an: 
J FIXED WiDRK SURFACE - t'ith adjustable chair and footres-
r FIXED CHAIR - with adjustable !i)0'"-ksur̂ "ace anc fvitrest 

STHTIO! 

106 

:. f,-, •-, 

BV -
k 
k 
k 
k 
k 
k 
k 
k 
k 
k 
k 
k 

*> e i 

liiith no adjustabie_feat:r^s 
'-^iiart one <^^ th(» t̂'?:'••'wing: -Fr-ess ''"2' 
" ' PJLLY-ArJUSTABLE ' 

>'IXcD-wORK-SURFACE 
-iXtD-'-nAiH 
STATICNARY 

; l i • li r *• •« A f: 

Down CF Unkri'jjn 
F4 r2 F3 

Done — Wiy 
FS F6 

10 w 
r5 -:U 

AVAy.v.vAVAy.yAyAv^Av.vAV,v/.y.v.vAVisy.yAyAv.vAv,y.vAy.y,sv.v,vj 

>,vAVAVAy.'iVAy.yi'A'.y,>wyiyj:»yAv,y.Nv.sv.vA^^^^^ 

Entir the '-'or-'k surface table tî 'iikr-i! 
Enter a oositive '•lumber. 

- i r . f ' T. u ••• • j .* " 

k ii 
r. ir. Mown CF u n n i j n Do^* — '''•'V HO Ĵ; -ii-' 

• ^ 2 =•: P i ~z -z - '• - 3 F'' 

,.̂ .,,,.,.,.,̂ v,.,̂ v,v,v,̂ y.̂ v.y.̂ ^ ŷ.y.̂ v̂.̂ y.y.sy.̂ v̂.̂ ^^^^^^^^^^^^^^ •̂̂  



107 

WORKFLACE-i is as fvl'ows: The ^'oi'owing I'.nclusions aiere '^a;* 
by '¥£ ERGONOMIST concerning the currint'Lorkolace Jesign. 

The workstation type is as ^o'lowss SI'/STAND 

Work surface upper ht. in lentiiTiet^rs is as f o l l o ' i s : 115.4 

Work surface 'ower !-it. in centirteters is as follows: NONE 

f^ C'-'air uooe''- height in centimeters is as *v;lows: iOl.c 

Chair lo'der height in centinieters is as fo::0'js: '3.0 

Foot'est ijppe-- nei'int in 'isntiTieters is as ^OMCWSJ :"*.? 

"ootr-ist ioujer heioht in ^lentimeters is as fol^oiis? 0 

M.ir'i»ai i - « - i r a ! icai reacr- in 'lentiiTiete's i; 1 5 -0l]0ii!S5 \52.-' 

I^aximum vertical reacri in centimeters is as foiiowsJ 205.4 

Normal horizontal reach in centimeters is as fjllows: 15.5 

^'aximum horizontal reach in centimeters is as follows: *:.'. 

"hi approximate reduction to the population that is to :-i 
accommodated at this workglace, due to a thigh clearance 
orobiem is as ̂ 'oTows: y'jOl 



APPENDIX D 

DEFINITION OF ANTHROPOMETRIC MEASUREMENTS 

(A) Shoulder Height - the vertical distance from the floor 
to the upper most point on the lateral edge of the 
shoulder with the operator standing erect. 

(B) Sitting Shoulder Height - the vertical distance from the 
sitting surface to the upper most point on the lateral 
edge of the shoulder with the operator sitting erect. 

(C) Body Depth - the maximum horizontal distance between the 
vertical planes passing through the most anterior and 
posterior points on the trunk. 

(D) Thigh Clearance - the vertical distance from the sitting 
surface to the top of the thigh at its intersection with 
the abdomen. 

(E) Forearm Length - the horizontal distance from the tip of 
the elbow to the tip of the longest finger. 

(F) Arm Reach - the horizontal distance from the posterior 
surface of the shoulder to the tip of the extended 
middle finger. 

(G) Elbow Height - the vertical distance from the floor to 
the depression at the elbow between the bones of the 
upper arm and forearm. 

(H) Sitting Elbow Height - the vertical distance from the 
sitting surface to the bottom of the elbow. 

(I) Popliteal Height - the vertical distance from the floor 
to the underside of the thigh immediately behind the 
knee. 

Refer to Figure 4 for a sketch of these measurements. 

108 



109 

FIGURE 4. Sketch of Anthropometric Measurements