Professional Documents
Culture Documents
DATA FOR
EQUIPMENT DESIGN
NATO CONFERENCE SERIES
I Ecology
II Systems Science
III Human Factors
IV Marine Sciences
V Air-Sea Interactions
VI Materials Science
AI rights reserved
Heinz Schmidtke
v
CONTENTS
vii
viii CONTENTS
Heinz Schmidtke
The past few years have seen efforts on the part of national as
well as international institutions to develop ergonomical standards.
These efforts underly various philosophies. In the Federal Republic
of Germany the prevalent view is to distinguish between appliances
or object-related standards, on the one hand, and basic standards,
on the other hand. According to this view, ergonomical standards can
only be categorized under basic standards. The expert institutions,
dealing with the creation of standards, consequently take great care
not to follow a collision course with the standards of equipment
but rather to advise them thoroughly. The validity of the data basis
of ergonomic standards constitutes the main problem area, notwith-
standing the kind of philosophy applied, on which the task of stand-
ardization is based. It requires only little imagination to suspect
that users of pseudo-valid data undertake everything possible to
convert their data records into standards, particularly under such
circumstances where this data has already been used in the manufact-
ure of equipment. In view of the fact that a modification of basic
standards will necessarily lead to a redesign of equipment and tools
it is no wonder to find manufacturers defending their data basis ve-
hemently.
switches and press keys are required for equipment fulfilling a si-
milar function, ARPA-radar even requiring controls in excess there-
of. This development is not only less beneficial to users but, in ad-
dition, promotes control errors with possible critical, security-
orientated consequences.
Bernhard DOring
Wachtberg-Werthhoven/FRG
INTRODUCTION
15
16 B. DORING
Environment
SYSTEMS DESIGN
Other
Ph,.ical ANALYTICAL
Science. TECHNIQUES
To all To all
Ph,.ical Scienc .. Hu.an Science.
- development phase
o concept
o definition
o design
- realisation phase
o production
o introduction
- operational phase.
------
> __
------
~"- TERMINATE
STUDIES ...z
w
I. ____ _ ~
w
I <!)
«
I z
«
...u
~
-- w
.....
o
0::
a..
REALISATION
------
PHASE
------
OPERATIONAL
PHASE
ever, in both cases the system life cycle just described starts again
with a new development phase in which new problem situations and
technical solution principles have to be considered. The system ac-
quisition process described is controlled bythe.project management
indicated by the box on the right side of Fig. 3.
- goal elaboration,
- solution generation, and
- solution selection.
GOAL
ELABORATION
SOLUTION
GENERATION
SOLUTION
SELECTION
function allocation {
man - machine
l
ma'::I:"!'=~ {
estimation
1
5~~~~:!~ { ~ ...
interfae.s ~
1
workstation
{ (operOtor"""
______ --' ...... mOc:hlM-]
IL _____ _
T
INTERFACE
TEM~TURE
,
HUMIDITY ILLUMINATION
\ , ;.,NOISE
_k
environment (,.-opera....:.-:t-or~) ~ I-m"a-'ch~ine-"
T
'\
TOXIC
. . . RADIATION
VIBRATION SUBSTANCES
"I machine II
"I machine 21
.. I machine 3 1
personnel selection
and training
Basic to this design task are, again, the behavioral task re-
quirementsestablished. Because not only must the critical values
of environmental factors recommended in the relevant literature be
considered but also with some factors task specific requirements
have to be specified and met for effective task performance by per-
sonnel.
CLOSING REMARKS
~
.Work Station Design
.Work Environment Design
• Workload Evaluation
• Persorv.! Selection and
Training
REFERENCES
Heiner Bubb
Mlinchen/FRG
31
32 H.BUBB
/'"
envi ronment
task =
{man machine}----. result
~~------------------------------~~
But first let us replace the second man in the block diagram
of Fig. 2 by a machine, especially a drilling machine. Of course
input and output of the man remains the same. At the input side of
the machine there are the controls, by which the operator "tells"
34 H.BUBB
speaker listener
2!
'"c:0 c:
'"E
-[
~
task ~
E result
r---- Q;'"
:::>
0 man u machine r--r-
'"'"c: '"E
:::> e
c:
~
'"'" 0
u
Operation
1 N
N Ln=1 ) • 100 %. (1)
no no
no
Manner of control
1iBBB lL
,
100' 90'
80'
4000 I'I" 1 II Ii
40· 45' Ii
2000 manCl!uvrable aero
20'
1000 ~ ~
800
...
C
10'
8' 400 I' Ii
...> I
I'
E
4'
200
Ii
0
E
'0 2' ~
... 19B
=
"0
l'
40 I' I!
I!
Q.
E
0
20 ~
10 ~ I!
8
I' Li
I' ~ Ii
li
81 81 81 81 8
10-4 10-3 10- 1 1 Hz
manCl!uvroble cutoff - frequency
Fig. 9. Lower and upper cutoff-frequency of man
w
max
p
.
x
w
max
max
respectively: ;;-- (2)
max
w
max
respectively: =-;r-
max
In most cases tasks are not given in such narrow bounds as the
dynamic task. Most tasks are characterized by a time independent
instruction related to the desired final product. We call that
static tasks (see Fig. 10). An example is drilling with a machine
or working with a turning-lathe. Only the final properties are pre-
scribed by the blueprint and not the way to them.
- - r-- r--
~
0 posi-
.e
~
i- i- r'- -
may be derivated. These dates represent the task of the next lower
level, the guiding task. The exact way and the necessary changes of
velocity or altitude are derived depending on the momentary circum-
stances. They represent the task of the lowest level, the stabilizing
task. These demands are realized by the control system, i.e. the
vehicle. In the case of a motor car that means that on the level of
the guiding task the driver places an individual course on the road
ahead depending on the bounds of the road, road conditions, other
vehicles and pedestrians, and he intends to keep the course on the
level of the stabilizing task. As you see this lowest level now is
a dynamic task.
Dimensionality
>
.Manner of task
task resul t
------------~-( man machine l
task result
-----------I-~-I man
I
closed man - machl ne loop
But also some human properties are important for the decision
man or machine: the control and supervisory task of a monitive
system are characterized by monotony. And monitive systems rather
require good abilities in decision making. In contrast to that,
active systems stress the man by the specific pressure to permanent
attention. Active systems require a high level of practice. Even by
the loss of the level of practice there arise difficulties, when
switching from automatic to manual operation (Edwards, 1976; Johann-
sen, 1976).
Display
task t man ).
j
machi ne
resul t
PurSUI lOS
(ompfnsolory los
task
result
task
+-c display ~ machine If result
congruent with the signal of the aeroplane in order to fix its po-
sition. Therefore he is able to observe separately the motion of
the forcing function, i.e. the task respectively the aeroplane, and
the motion of the result, i.e. the position of the cursor. He may
get experience of the motion of task and result independent of each
other. Thus he is able to make a good short time prediction of the
future motion and so to react in time. But by pursuit displaying
the whole area of variation of task and result is to be displayed.
That means, that possibly both appear very small.
task resu II
real situation
F
~
•
simulated situation 'R-~
R F
structur. diagram
compatible situation
F
•
real situation
(~
IIY lng Wllh exlernal sighl flyong wllhaul pxlernal Sighl IIYlng wllhaul pxlernal Sighl
ccmpaloble SIIuol,on
The aim of this chapter was to show, how machines might be ad-
apted to the properties of the man. Therefore the performance of
the system may be improved and simultaneously the load on the man
be decreased. This becomes possible, if you do not start from the
possibilities of technology but start from the today's knowledge of
information processing by man. By such a process man becomes able
again to master the machine, even under the condition of highest
complexity, and for that reason modern apprehensions that man could
be controlled by technology, can be prevented.
REFERENCES
R. Bernotat
Wachtberg-Werthoven/FRG
INTRODUCTION
57
58 R. BERNOTAT
o
Application of Ergonomics sometimes delays the design
process, at least if it comes into the design process
late.
o
It can often not be shown what the quantitative benefit
in selling the product will be.
.J: It.~]
U
ao
- -... Imachine!
27"1
192H
relationships
8:a ~
u
~
]a ;-~~-~~~-------------------------------------------
partial models
of
human operator
experiments
Isubjects 1.....-:.....l"!U!'!!'!_~
.J: if data and models
U
ao
. ---+ ~------,
I machine •
are insufficient
Q. or
g.
if aiming at
S optimal solution
i
E
";j
..
~
Types of experiment
field experiment
laboratory experiment
performance
to the human operator by ergonomic means. Some of the data are given
in the form of numbers or tables, but most data are presented in
diagrams A = feB) etc.
Measuring problems
Factorial Experiments
----------------,
I
HUMAN OPERATOR
I
Visual scanning
I
I Audio discrimination I
I I
l
Learning rote Probability of Hror
I Physiological chonges Time to respond
I
Environlnent I Muscle e.ertion I ResponH consistency
I I RnponH rang.
~ I Control movement
rI
Proc~ur.sl fr~u.ncy and I R.sponH accuracy
I amplituM I
I
I
~F :
I INPUT LOAD ,:
I
I
I
-:OPERATOR EFFORT:
I
I
I
I
-, PERFORMANCE
INDEPENCENT 'tMIABlE
j
I IUotBER AN) TYPE
j
EXTRANEOUS ~BlES -----t.~ SUBJECTS -+----OF SlBJECTS
• ELIMINATE _ I REPRESENTATIVE SAMPlE)
• HOlD CONSTANT
• COMPENSATE
MEASURNENTS.OEPEfClENT VARIABlES
- SUBJECTIVE
- OBJECTIVE
independent variables
\\1//__
Extraneous variables _ _ _.....1 subject ..,I~ ~_ n subjects
• • (representative sample I
11\
dependent variables
Number of subjects
Types of subjects
A much more serious effect may occur if subjects used are not
"representative", that means that they are not typical for the
future user.
68 A. BERNOTAT
Ix;plctancy 11
standard deviation C1
pix)
population
normal distribution
11-0 11+0
assumed values
for
mean and S.D.
sample
d
assumed distribution
10
B
.B performance degradation population
.;: due to fatigue end/or performance
"til stress distribution
:.0 , .... - . . ,
of users
r:-
~ I" "
o I
.c
[ ~
I
)(
sample
performance of
subjects abcNe average
e.g. RI D staff
Learning time
error ~ subject 2
subject 3
I
A
perf-l
statIonary J
l'
-
'",.,,,
2 4 6 8 10 12 14 16 18 20
NPUcatilllll of....,....
~
.....
\ear...... time "I
Fig. 9. Learning Curves and MMS-Performance
• COLLECTION OF DATA
-data Ax A
-diagrams
d~
-distri but ions
Ax
• ANALYSIS
determination of el""ents
and their relationship
• SYNTHESIS
integration of knowledge
about
influences
al the human ,l?roperties
and capabilities in form
J
of partial models
• DESIGN THEORIES
de'lelopment of higher order design-theories based
on the synthesis-models leading to ergonomic design
procedures for man-machine-systems which are
teachable and applicable.
Evaluation methods are an important part of these theories
Research
o Measuring methods have to be improved especially concerning
mental workload and influences from the social environment.
o
Some agreement or even "standardization" of basic measuring
procedures would help to make data comparable.
o
Moxomultifactorial experiments should be started to get per-
formance predictions closer to real World performance. How-
ever, because such research requires long term efforts,
funding may be difficult.
o Software Ergonomics with increasing automation and computer
application becomes an increasingly interesting and impor-
tant topic.
o
More research should be directed towards the assessment of
human failures, their type and probability, so that such
GENERATION OF ERGONOMIC DATA 73
guidance
4----1 .rror
pion execution
system
Application
REFERENCES
Roebuck, J., Kraemer, K., and Thomson, W., 1975, "Engineering Anthro-
pometry Methods," Wiley, New York (N. Y.).
Rouse, W., 1981, Human-computer interaction in the control of dynamic
systems, Computing Surveys, 13:72.
Rutenfranz, J., Luczak, H., Lehnert, G., Rohmert, W., and
Szadowski, D., 1980, "Denkschrift zur Lage der Arbeitsmedizin
und der Ergonomie in der Bundesrepublik Deutschland," Boldt,
Boppard.
Seifert, D., and Doering, B., 1981, SAINT - Ein Verfahren zur Model-
lierung, Simulation und Analyse von Mensch-Maschine-Systemen,
z. f. Angewandte Systemanalyse, 2:127.
Shackel, B., ed., 1973, "International Symposium on Ergonomics and
Standards," Proceedings University of Technology, Loughborough.
Sheridan, T., and Ferrell, W., 1974, "Man-Machine Systems: Infor-
mation, Control and Decision Models of Human Performance,"
MIT Press, Cambridge (Mass.).
Spillers, W., 1974, "Basic Questions of Design Theory," North-Holland,
Amsterdam.
SWain, A., and Guttman, H., 1980, Handbook of Human Reliability Ana-
lysis with Emphasis on Nuclear Power Plant Applications,
Sandia Laboratories, Albuquerque, prepared for Office of
Nuclear Regulatory Research, Washington, D. C., NRC FIN, No.
A 1188.
Wortman, D., Duket, S., and Seifert, D., 1977, Modeling and Analysis
using SAINT, a combined discrete/continuous Network Simula-
tion Language, in: Proceedings, Winter Simulation Conference,
U.S. Nat. Bur. of Standards.
SOME LIMITATIONS TO ERGONOMIC DESIGN
J. Moraal
INTRODUcrION
77
-.J
(X)
Table 1. Performance Shaping Factors
-.oJ
CD
80 J.MORAAL
ing tasks. The recipe is simple. Firstly, find out what has to be
done by an analysis of the aims of a system and of the task de-
mands. Secondly, find out what man can do, i.e. investigate his per-
formance capabilities. Thirdly, fill the gap between both sets of
data by choosing the right design of equipment and environment and
by the right choice of task aids. Given, that man or the human op-
erator is well-selected, well-trained and well-motivated, the final
result will be optimal task performance. However, this recipe is too
simple. Man is not a given entity, always operating in the same way,
stable under different conditions of the environment, unaffected by
personal and emotional matters and with predictable performance. In-
stead, human performance is difficult, sometimes very difficult to
predict. Apart from what the human engineer wants that the operator
brings with him in the task situation, like his skills, experience
etc., he always brings more: his motivations, beliefs, wishes, ex-
pections and, in particular, his variability.
ANTHROPOMETRIC VARIABILITY
After normalizing the data from Tables 3 and 4 for length and
weight, Haselgrave was able to extract other factors influencing
body shape, namely, torso depth, limb length and torso width, to-
gether accounting for about 33% of the variance in the male and
female samples. These factors then, seem to represent main features
of body shapes.
PERFORMANCE VARIABILITY
(XI
Co)
(Xl
~
Weight, kg 416 62'1 9·52 4J8 49·8 51·8 54·4 56·6 58·5 61'1 63·0 65'6 68·6 74·3 79·9 90·7
Stature, mm 416 1624·5 56·01 1493 1537 1553 1578 1594 1609 1623 1637 1652 1672 1699 1719 1763
Eye height, mm 416 759·7 29·03 689 707 722 735 746 754 762 769 775 784 795 805 821
Shoulder height, mm 416 580·2 25·00 526 540 548 557 567 574 580 587 593 602 614 623 635
Neck width, mm 415 129·7 13042 97 108 112 120 123 126 130 133 137 140 147 152 160
Shoulder width, mm 416 339·4 22·68 289 304 313 320 327 334 338 344 350 356 368 381 398
Chest breadth, mm 416 281·8 21·55 240 250 258 265 270 274 279 284 291 297 311 319 350
Seat breadth, mm 416 364·0 26·92 310 325 333 342 349 356 362 369 376 385 399 412 437
Chest depth, mm 415 251·7 28·34 203 212 220 228 234 239 247 255 264 273 291 305 333
Stomach depth, mm 416 253-3 34·11 197 207 215 224 233 239 248 258 267 278 297 320 359
Thigh depth, mm 416 156·0 15·66 126 134- 138 143 147 151 155 159 162 167 176 181 204
Knee height, mm 415 540·6 26-73 479 499 507 518 526 534 540 547 553 562 575 588 606
Buttock-knee
length, mm 415 600·9 26·89 540 560 567 577 586 594 601 608 616 622 636 646 666
Arm length, mm 257 721·8 30·34 656 672 681 698 707 715 724 729 736 742 760 771 801
Shoulder-elbow
length, mm 416 351·4 15·82 316 327 330 338 343 347 351 355 360 365 371 378 393
Elbow-fingertip
length, mm 416 426·8 18·32 382 397 403 412 418 422 426 432 436 442 450 456 473
Shoulder slope,
u
degrees 413 22·2 3·77 12·8 15·8 17·2 19-4 20·3 21-2 22·2 23-2 - 24·6 25-7' 27·0' 28T 30·l
~
s:
0
:tl
»
»
r
SOME LIMITATIONS TO ERGONOMIC DESIGN 85
.'\
o \/\ (48.3km/h)
. •
-; 0.8
~0.6
:c
.8 0.4
ea. 0.2
CONFLICTING CRITERIA
FUTURE RESEARCH
CD
~
92 J. MORAAL
REFERENCES
Hans W. Jurgens
Kiel/FRG
93
94 H. W. JORGENS
This seems quite clear at first glance, but Table 1 shows the
very considerable variation in values for three particular body mea-
surements, namely height, ankle-knee length and buttocks width, if
the percentile groups are differentiated for example according to
age and sex. The range of variation that then becomes apparent may
be as much as 41 cm in height or 13.8 cm in buttocks width. Trying
to take just this variation into account leads to technical diffi-
culties; it is probably not possible for a young man of the p5 group
for buttocks width to rest his arms on the arm rests of a chair that
suits the requirements of an elderly lady of the p95 group for the
same parameter. Similar examples will be found for other measure-
ments.
Japan 67 %
Fed.Rep.Ger. 65 %
U.K. 63 %
China 63 %
Eire 58 %
USA 57 %
Turkey 56 %
India 56 %
the variations at the two ends of the scale are so extreme that ac-
counting for these minorities would frequently lead to complications
or even dangerous situations for the large majority otherwise c~ter
ed for. This means the jobs must in some cases exclude particularly
small or particularly tall people.
p 5 194.8 cm p 95 224.8 cm
p 1 183.7 cm p 99 232.4 cm
Difference 11.1 cm 7.6 cm
Differences combined 18.7 cm
SPECIAL PROBLEMS
troops to those not above the 30th height percentile leaves enough
men for tanks, we must recall that the number available will be
halved in the course of the next ten years. Therefore the question
must then be reconsidered, whether the 30th percentile for height
must still be regarded as the upper limit. Added to this, the
secular acceleration trend is still leading to a constant, if
small,increase in height, so that it is not possible to limit
indefinitely the space available for the person carrying out a
given job. It must be generally accepted that the ergonomist
cannot only think in terms of technical ergonometric factors, but
that if he is basing his calculations on the requirements of a
population of humans, he must always relate his findings to the
population in question.
REFERENCES
Karl H. E. Kroemer
Blacksburg, Virginia/USA
PREFACE
101
102 K. H. E. KROEMER
Looking at medieval body armor one cannot help but notice that
today's males would have a hard time fitting into these small
shells. Such secular increase in body size is also apparant from the
everyday experience of children being larger than their parents.
While evidence for long-term gains in body size is only indirect,
statistical comparison of anthropometric data taken by the military
services on large samples with consistent techniques proves in-
creases in stature in the neighborhood of one centimeter per decade
throughout much of this century. In the U.S.A., a gain in body weight
of about 2 kg per decade has become apparent during the second half
of this century.
SURVEY SAMPLING
The more specialized the sample and the dimensions taken, the
more this method resembles "subgrouping", such as applied to pre-
selected and pre-measured individuals usually representing the ex-
treme end of the population distributions, e.g., the tall-heavy,
or short-light. Such subgroups are often used by engineers to check
their designs for fit to the extremes of the user population, as-
suming that the mid-ranges will be accommodated if the design fits
the unusual persons. The ultimate is, of course, measuring indivi-
duals in order to fit equipment exactly to the person, such as in
tailoring space suits for selected astronauts.
104 K. H. E. KROEMER
MEASUREMENTS NEEDED
The first question is: Who needs anthropometric data for what
purpose? A manufacturer of loosely-fitting garments is interested in
different anthropometric data than a physical anthropologist. The
manufacturer of face masks needs very different information than an
automobile designer. To determine the size and location of safe
openings in equipment so that the operator cannot be caught in them
requires information other than needed to determine leg room, and
chair adjustments, for workstations used by seated operators.
MEASUREMENT TECHNIQUES
Photography does not need physical contact, but can relate only
to the surface contours of the body which are, with few exceptions,
determined by compressible tissue and give little information about
the underlying bony landmarks preferred by the anthropologist and by
the engineer. Standard photography, with its techniques derived from
cartography, provides 3-D pictures of the surface contours of the
human body. To overcome the difficulty of relating surface features
to underlying bony structures, the use of "point marks" has been
proposed (Reynolds, 1982) which would allow locating limbs (re-
presented by the point marks) with respect to each other in spatial
coordinates" This technique has yet to be applied to living subjects
in general surveys.
Table 1. u.s. Civilian Body Dimensions, Female/Male in cm or kg, for Ages 20 to 60 Years*
Percentiles
5th 50th 95th Std. Dev.
Head Breadth 13.6/ 14.4 14.5 / 15.4 15.5 / 16.4 .57/ .59
Head Circumference 52.2/ 53.8 54.9 / 56.8 57.7/59.3 1.63/1.68 ?"
::t
Interpupillary Distance 5.1/ 5.5 5.83/ 6.20 6.5 / 6.8 .44/ .39
!T1
~
;:g
0
m
3:
m
;:g
m
Z
G)
Z
m
Percentiles m
:lJ
5th 50th 95th Std. Dev. Z
G)
~
Forward Reach, functional 64.0/ 76.3 71.0 / 82.5 79.0 / 88.3 4.5 / 3.6** Z
-I
Elbow-Fingertip Length 38.5/ 44.1 42.1 / 47.9 46.0 / 51. 4 2.2 / 2.2** J:
Hand Length 16.4/ 17.6 17.95/ 19.05 :lJ
19.8 / 20.6 1.04/ .93 0
Hand Breadth, Metacarpale 7.0/ 8.2 7.66/ 8.88 8.4 / 9.8 .41/ .47 "'0
0
Hand Circumference, 3:
Metacarpale 16.9/ 19.9 18.36/ 21.55 19.9 / 23.5 .80/ 1.09 m
-I
:lJ
-<
Chest Depth 21.4/ 21.4 24.2 / 24.2 29.7/27.6 2.5 / 1.9**
Elbow-to-Elbow Breadth 31.5/ 35.0 38.4 / 41. 7 49.1 / 50.6 5.4 / 4.6
Hip Breadth, sitting 31.2/ 30.8 36.4 / 35.4 43.7/40.6 3.7/2.8
Buttock-Knee Length, sitting 51.8/ 54.0 56.9 / 59.4 62.5 / 64.2 3.1 / 3.0
Foot Length 22.3/ 24.8 24.1 / 26.9 26.2 / 29.0 1.19/ 1.28
Foot Breadth 8.1/ 9.0 8.84/ 9.79 9.7/10.7 .50/ .53
Weight (in kg) 46.2/ 56.2 61.1 / 74.0 89.9 / 97.1 13.8 /12.6
...o
'-I
108 K.H.E.KROEMER
Stature and weight are the best single predictors of all other
body dimensions. This model has been checked recently by Robinette
and Churchill (1979) and found to be fully valid. Using regression
equations which need as input variables body height and weight, equa-
tions have been developed that predict other body dimensions with
high accuracy (Robinette and McConville, 1981). Body dimensions cal-
culated through regression equations are additive.
The eye. Called eye design point, eye ellipse, line of v~s~on,
etc., in different industries, this point constitutes the geo-
metrical link between targets and the eyes of the operator, usually
assumed to be seated.
The hip. Depending on industry usage, either an approximation
of the hip joint is used for design purposes (H-point), or the in-
tersection of the seat pan and the backrest planes (Seat Reference
Point, SRP) is used to establish the link with the body of the
seated operator.
The foot. Called heel-rest point, or accelerator-heel point, or
package-origin point, this reference location connects one or both
feet of the sitting operator with the equipment.
The hand. Despite its importance, and because of the diffi-
culties in defining specialized activities, this interaction between
the human body and the equipment is only loosely described such as
reach envelope, as preferred manipulation area, or as work plane
projected onto horizontal bench or table surfaces.
While all of these interface points are important, the inter-
face between the hand and the object is usually critical. However,
it has not been used to the extent desirable because of the diffi-
culty of establishing a hand or finger "point". The body system con-
necting the fingers with the main part of the body is of· such vary-
ing dimensions, of such high variability in its positions, and in-
fluenced by so many external features (design of the work place, of
the object, of the task, etc.) that in the past attempts to link the
operator with the work station through the hands proved to be too
difficult.
u.s. ARMY
The u.s. Army has not only collected extensive data on its own
personnel, but also obtained the body dimensions of various foreign
populations. (DOD-USA-DRDNA-ES, Natick, Massachusetts 01760)
N~A
ERGODATA
REFERENCES
Heinz-Peter RUhmann
MUnchen/FRG
INTRODUCTION
115
116 H.-P. ROHMANN
Basic data for the design of work places result from the fea-
turesof the human visual system (lines of sight, visual fields),
from the sitting positions mainly taken up at the work place (e.g.
rearward seating position for predOminantly monitoring a system)
and from the dimensions of the hand reach area.
Lines of sight
If the head and the eyes are kept in a straight (tense) position
the line of sight is identical with the horizontal (=0°).
Fields of view
fVi/_\:::::::---oo
Horizontal line of sight 1--10°-15°
rel ated to the head
kI /~k- -- --
,~'05-"OO
i? ,",-::::-
Nonnal line of sight 0°
10°-15°
Head and eyes ina
rel axed position II '
'25°-35° Nonnalline
of sight
when the head and the eyes are reposing. One distinguishes between
the monocular and the binocular field of view.
defined above one obtains the secondary and tertiary field of view
(monocular and binocular).
Sitting positions
D
.~
. . . . . . . . tUICl U_l • • 1
•••• ,., c.l •• ~ .,t_ "
o ,
(~J
0.1, •• 1 •• ,.el 'l_l •••
• , •• f ... HI t _'t •• 1t
Hand reach
The maximum hand reach capability and the hand control reach
may be defined as follows.
11.141 .f
.t,. Ii
V'IOm-._....nIl'>;w , 0 d
,~""~:-..:htj..". III 1lI./ • II ow
o •
•• l •••
•• f.I
4.,.cI ".'4 of
c.10., .". it
Basic data for the determination of the hand control reach are
the body dimensions of men belonging to the 5th percentile level.
Data are available for standing and sitting position.
BASIC OAT A FOR THE DESIGN OF CONSOLES 121
.5° 0°
00
~
0
1111. 110" ~ •• ,.u • •• condery
Vi.W for l'lh, el,.ut,
f,.14 of
~~
~.
.,u.
rod
. bluol,oll" ..
Thus, the side view contour of the maximum hand control reach
in a plane parallel to the sagittal plane is designed by turning
the arm around the shoulder pivot (Fig. 9). For this procp.dure the
elbow angle is set to 180°.
BASIC DATA FOR THE DESIGN OF CONSOLES 123
100·
-if--r-+---\-++-+-OO
~rtl~----+o·
r .. " 01
: ",.v for colo~r .11 ,.ul,
For the time being, a movable drafting template for the design
of the plan view of the human body and the hand control reach is not
available. However, the hand control reach can be defined construct-
ively by approximation in a simple way.
,
Fig. 8. Basic data resulting from an upright (normal) and from a
rearward sitting position
ri.ght sitting position. The vertical distance between the seat and
the surface below the work surface is derivable from the thigh clear-
ance height of the 5th percentile-mrueand a vertical clearance bet-
ween the thigh and the surface below the work surface amounting to
20-30 mm. This procedure leads to a vertical distance between the
seat and the surface below the work surface of 150 mm.
Console
/
/
/
/
I
rt~t.1...
I
I
I ~~
"'J- . . . . . sOP
\I
1·I \ .-;i
I
I "-
\ For the 5th percentile·.. ":
\
\
\
"-
"-,
......
Fig. 10. Construction of the plan view contour of the hand control
reach
The front part of the work surface is usable for writing work,
for the setting up of equipment (e.g. keyboards, Fig. 13 bottom) ,
or for the installation of indicators and controls (Fig. 13 top).
Viewing the consoles from the top, they can have a straight
or segmented design.
Information board
cess by direct view. T,he lower window frame of the objects of in-
terest must be above the line limiting the view ahead.
Panel
Depending on the type of the control room, not only the spatial
arrangem~nt of the elements defined above is specified, but also
their interrelation.
Fig. 15. Panel with a work surface inclined towards the operator
DIMENSIONS
"
~
Fig. 16. Arrangement principles of the work stations according to different types of control ~
rooms ~
o
~
~
~
Z
Z
BASIC DATA FOR THE DESIGN OF CONSOLES 133
Clearance dimensions
Leg clearance
For the leg space of consoles with work surfaces where displays
and controls are to be ~nstalled, the dimensions of Fig. 18 are re-
commended. In view of these dimensions the 95th percentile-male will
find acceptable seating conditions.
Fig. 17. Leg clearance at a console with a work surface for writing
or for the setting-up of equipment (no equipment installed
into the work surface)
For the lay-out of the visual field and for the determination
of the blind sectors, a design eye point has to be established.
Consoles
~
»
a
:D
-t
J:
m
C
m
en
C5
z
o
."
n
oz
en
or-
m
en
Fig. 20. Basic data for consoles with and without an upper segment
I
T 500 max. 790 max. w
-- ---------- _ ..- -----------
....
138 H.-P. ROHMANN
Fig. 22. Basic data for consoles with vertical stacked segments
Information board
the visual distance and the back\olard inclination angle (y) of the
mounting surface for displays. These data result from the geometry
of sight related to the design eye point and further data represented
in Fig. 23.
Example:
HAP = 1210 mm, LAP = 150 mm, HSV = 910 mm, LSV = 850 mm
Then:
y = a - B/2 = 1 3 • 3 ° .
Fig. 24. Basic data for the lay-out of the foot rest
male. That is ensured if the depth is at least 330 mm. If foot rests
are fixed to the floor the front edge of the step should be 150 mm
behind the vertical reference' line. For this kind of lay-out a
longitudinal adjustment is not necessary. The horizontal extension
of the step should be at least 500 mm. A foot-operated control pro-
vides for vertical adjustment and, if available, for the adjustment
of the inclination, too.
CONCLUDING REMARKS
In Fig. 25 three consoles are drawn one above the other. This
type of console is required for a sitting position with vision over
the top border. It is furthermore assumend that this type of console
is mainly used for manual process control from an upright (normal)
seated position.
Fig. 25. Consoles for sitting position with vision over top
standard console middle-sized (MIL-STD-1472B)
- - - - - . console according to a working draft (DIN)
type: equipment installed into the desk sur-
face, predominantly manual process
control
-.-.-.-.-.-: console according to a French draft proposal
("Positions and dimensions of men working on
machines and plant")
REFERENCES
Damon, A., Stoudt, H. W., and McFarland, R. A., 1966, "The Human
Body in Equipment Design," Harvard University Press, Cambridge
(Mass.).
Schmidtke, H., 1981, Arbeitsplatzgestaltung, in: "Lehrbuch der Ergo-
nomie ," H. Schmidtke, ed., Hanser, Mlinchen, Wien.
Schmidtke, H., and Ruhmann, H. P., 1981, Betriebsmittelgestaltung,
in: "Lehrbuch der Ergonomie," H. Schmidtke, ed., Hanser,
Munchen, Wien.
144 H.-P. RUHMANN
William H. Cushman
INTRODUCTION
145
146 W. H. CUSHMAN
FUNCTIONAL ANTHROPOMETRY
of about 7 1/2 inches (19 cm) between the resting position of the
elbows and the front edge of the work surface.
The Chaffee and Emanuel (1964) data may be used to help decide
where to place controls on vertical surfaces, such as instrument
racks, if the operators are to remain seated. The data also suggested
that the depth of a console shelf should not exceed 20 inches (51 cm)
and that 28 inches (71 cm) is the preferred shelf height. The latter
recommendation is one inch below the height recommended by Ayoub
(1973). (See section on Work Surface Height for additional details.)
VISUAL CONSIDERATIONS
MarOR PERFORMANCE
sagittal plane and 25 inches (64 cm) above the SRP. The fast per-
formance area for the toggle switch was considerably smaller than
the corresponding areas for both the knob and push button. This
finding suggests that the selection of a location for a toggle
switch is more critical than the selection of a location for the
other two types of controls.
~~~----------~
~\{,~
~,(\.
~"3
Fig. 1. Console arrangement for the Siegel and Brown (1958) study.
front panel and two side panels. Number of seat movements, magnitudes
of seat displacement, number and extent of body movements, and number
of partial and full arm extensions were recorded. Subjective evalua-
tions and preference rankings were also obtained.
Dimensions in Inches*
Type of Console A B G H
Sit-stand 62 26 36 35
(157) (66) (91) (89)
Stand (with 62 26 36
vision over top) ( 157) (66) (91 )
Stand (without 72 36 36
vision over top) ( 183) (91 ) (91 )
*Equivalent dimensions in em are given in parentheses.
anA" must never be more than 29~ in. (75 em) greater than J.
b"A" must never be more than 33~ in. (85 em) greater than J.
1
A
measured from "the seat back to the front of the knees" (Ayoub and
Halcomb, 1976, p. B-55).
CONCLUDING REMARKS
Design methods are also important, but they are beyond the
scope of the topics to be discussed at this NATO Advanced Research
Institute. Nevertheless, designers should be familiar with at least
some of the powerful design tools that are available. These tools
include procedures for determining panel layouts (e.g., link analysis
and linear programming), computer-aided design, mathematical model-:-
ing, and mock-up construction and evaluation.
When using ergonomic data, the designer must pay special atten-
tion to the conditions under which the data were obtained and apply
correction factors if needed. It was pointed out earlier that factors
such as body dimensions, hand orientation, seat characteristics, type
of restraint, and type of control have a significant effect on reach
envelopes. Application of reach data without making allowances for
these variables may lead to an unsatisfactory design. The amount of
effort that should be spent in modifying existing data so that they
may be used for new applications will depend upon the criticalness
of tasks that the operators perform and the consequences of operator
errors.
REFERENCES
Roth, J. T., Ayoub, M. M., and Halcomb, C. G., 1977, Seating, con-
sole and workplace design: Seated operator reach profiles,
Proceedings of the 21st Annual Meeting of the Human Factors
Society, 83.
Sharp, E. D., and Hornseth, J. P., 1965, The effects of control lo-
cation upon performance time for knob, toggle switch, and
push button, AMRLrTR-65-41.
Siegel, A. I., and Brown, F. R., 1958, An experimental study of
control console design, Ergonomics, 1:251.
Squires, P. C., 1956, The shape of the normal work area, U.S. Navy
Medical Research Laboratory, Report No. 275, New London
(Conn. ) .
Stoudt, H. W., 1978, Arm-leg reach and work,space layout, in: "Anthro-
pometric Source Book," Vol. I, Webb Associates, ed~ NASA
Reference Publication, 1024.
VanCott, H. P., and Kinkade, R. G., eds., 1972, "Human Engineering
Guide to Equipment Design," U.S. Government Printing Office,
Washington, D. C.
Webb Associates, eds., 1978, "Anthropometric Source Book," Vol.I;
"Anthropometry for Designers," Vol. II; "Handbook of Anthro-
pometric Data," Vol. III; "Annotated Bibliography of Anthro-
pometry," NASA Reference Publication, 1024.
Woodson, W. E., 1981, "Human Factors Design Handbook," McGraw-Hill,
New York.
ERGONOMIC DATA FOR THE DESIGN OF BODY SUPPORT
Heinz Schmidtke
INTRODUCTION
159
160 H. SCHMIDTKE
- the position of the eye point or the eye ellipse (in relation to
the field of vision),
- the Seat Index Point (SIP), according to ISO 5353, ISO 6682, and
SAE J 1163;
- the Seat Reference Point (SRP), according to ISO 3462, DIN 1857,
part 2, and ISO 4253, DIN 1857, part 1, DIN 24 091 E, DIN 1858;
Fig. 2. Device for the determination of the Seat Index Point (SIP)
• Small lateral upholstery or broad seat and backrest between the lateral
upholstery, hard Isoll upholstery
• Marked lateral upholstery, hard upholstery
4_----10mm -----t
10 mm tolerance area according to ISO 3462
..
KM-LokO: .. +.cBMW 5er .. :MW 3er (SF,W)
•
•
Volvo 244 Ford Granado
BMW 200, I
BMW- Test - seat lift-truck
(2) Sable
"tAudi100
•
--tH-~kt
Direction of movement
I
Fig. 4. Location of the Seat Index Points of 15 different automobile
seats relative to the H-Point
II
99 104
- Seat and back rest should have a cover providing certain lateral
guiding, with a sufficient water vapour permeability and an ability
to absorb water vapour.
a} Size of seat
The depth of the seat should be between 430 and 450 mm, the width
between 400 and 460 mtl.
b} Size of back-rest
The minimal height of the back-rest should be 500 mm above the
seat. The clearance between seat and the lower flange of the
back-rest should not exeed 150 mm. The width of the back-rest
should be adapted to the width of the seat, with a minimum width
of 400 mm and, under certain circumstances, devices for neck-
rests are necessary.
c} Back-rest inclination
It is conceivable that out of economic reasons a design without
ERGONOMIC DATA FOR THE DESIGN OF BODY SUPPORT 165
Fig. 6. Data for the design of seats attached to the floor (side
view)
Fig. 7. Data for the design of seats attached to the floor (front
view)
166 H, SCHMIDTKE
d) Seat adjustment
In order to meet the requirement of adjusting the eye point of
the 5 th percentile operator to a point not less than 1210 rom
above floor, the height of the seat should be made adjustable,
i.e. for male and/or female operators from 450 rom to 530 rom
(= 80 rom).
f) Arm-rest
Floor-fixed seats, in general, should be fitted with upward-fold-
ing armrrests. The height of arm-rests above seat should be 190 rom
to 230 rom. Armrrests will be accepted by users if their surface
dimension is about 50 rom multiplied by 300 rom (maximum 350 rom).
If the inclination of the back-rest is adjustable, it should be
possible to keep the arm-rest constantly in a horizontal posi-
tion.
'r~y
- SlhlOt on pr or 0 racesogn
~ Impo"meot compnred
~ with stondord soot
a) Size of seat
The depth of the usable part of the seat should not exceed 400 mm
lest small female operators encounter difficulties in making pro-
per contact with the chair-back. The width is irrelevant in so
far that the value will not remain under 400 Mm.
b) Size of chair-back
If the chair-back is rigidly mounted the ~n1mum height should be
450 mm and the width 350 Mm. The vertical shape of the chair-back
. should be concavely dished, particularly in the lumbar region.
>~ '!he height of the main lumbar support should be between 80 and
100 mm above the point of contact with the seat.
c) Chair i~clination
It should be possible to adjust the seat of work-chairs used for
different sitting postures between - 2° and + 5° in correspondence
to the horizontal plane (see Fig. 10).
d) Height adjustment th
A work-chair is usable both for the 5 percentile female as well
as for the 95 th percentile male, provided the adjustment range
goes from 530 mm (upper position of seat) to 380 mm (lower posi-
tion). Regarding work on assembly lines, it may be necessary to
consider the real height of the work bench. This may be achieved
in the following way:
-'.. 0°
c:rI
A '\
\ s~
l'
B _5°
B·B
F====+-L....L...£ 2. ) '
{~, 8
0\ . 0\
)50
- The guiding ways or the shaft for the vertical adjustment of the
seat should be designed in such a manner as to prevent injury to
the back of the operator as well as the entanglement of the op-
erator's clothing.
a) Height adjustment
A standing-support may be used by both a tall man and a small
woman provided the range of seat adjustment goes from 700 rom
above floor to 850 mm.
b) Stability
Sufficient tip over-stability will be achieved if the front and
back contact surfaces of the substructures has a width of not
less than 450 rom and a distance in the longitudinal axis of about
500 rom. The contact surface with the floor should be coated with
antislip material.
c) Seat design
A standing-support should have a saddle-shaped bucket seat with
a back-width not less than 350 rnm and a minimal depth of 250 rnm.
d) Seat inclination
By using a standing-support both feet will be in contact with
the floor. In order to minimize body pressure the seat should be
tilted forward in progressive steps ranging from 0° to - 5° and
to - 10°.
- if, in relation to the seat height, the feet fail to make suf-
ficient contact with the floor;
- foot-rests for use with more or' less out-stretched legs should
have an adjustable angle of incidencei
a) Height adjustment
If the clearance underneath the table board is about 700 mm, a
5 th percentile female will reach the average height of the eye
point (1210 mm) only by adjusting the seat height to 520 mm.
Because the length of the lower legs (the length of the shoes
including) of the 5th percentile female is about 395 mm, a foot-
rest of at least 125 mm height is needed. A 5 th percentile male
will reach the same sitting conditions by a 30 mm height of the
foot-rest. It may be concluded that the height adjustment should
range from close to 0 up to 125 mm.
c) Width of foot-rests
A width of the foot-rest, equivalent to the width of leg room,
should be aimed for. 400 mm to 500 mm should be the minimum.
d) Adjustment mechanism
In the design of the adjustment mechanism for height and incli-
nation, consideration should be given to the possibility of the
operator having undergone a foot operation.
50
-0 0
The neck-rest design should not impair the vision in all direc-
tions.
a) Height of neck-rest
For the neck-rest to be completely integrated into the back-rest
of the seat, the horizontal plane through upper edge of the back-
rest should be at least 775 mID above H-point (approximately 875 mID
above unloaded seat). This will meet the requirements of the 95 th
percentile male. For the purpose of safety aspects however, it is
necessary to take the anthropometric data of the 99 th percentile
male into acount. For the aforesaid reason the value of 775 mID
above H-point has to be extended to a minimum of 800 mID.
ERGONOMIC DATA FOR THE DESIGN OF BODY SUPPORT 177
b) Width of neck-rest
The neck-rest should have a minimum width of 200 mm in order to
prevent a head-slide off if the application of acceleration forces
is transversal. In addition, slide-off effects may be reduced by
shape (see Fig. 15).
d) Upholstery
In order to reduce head acceleration the structure of neck-rest
should be built up, the inner part with rigid expanded plastics
or steel-deformation elements, and the surface padded with soft
upholstery.
178 H. SCHMIDTKE
REFERENCES
Walter W. Wierwille
INTRODUCTION
This paper will also present two new areas associated with
control design. The purpose of including these areas is to indicate
how little is known about important particular classes of controls
179
180 W. W. WIERWILLE
and to show what future trends may take place in control design.
The problems are typical of those that may be encountered, but they
are by no means the only ones that will be encountered.
General Principles
Manual Control
Bang-bang Controls
b. Crane dynamics
The dynamics used in each axis of this two-axis study were as-
sumed to have the form
where Fist the constant force applied when one of the control but-
tons is depressed, x is the position of the output, M is the equi-
valent mass, and D is the equivalent viscous friction of the system.
The dynamics can be rewritten in transfer function form as
Xes) F/D
U(s) s
s (1 + D/M)
where U(s) ! for a step input, F/D is the system gain and D/M is
the system co?ner frequency in radians per second.
spot was within the target area for 1 second, the trial ended. Each
subject completed sixty-four trials, the first thirty-two of which
were practice.
Multi-function Controls
8
OJ
,/0
~ • 10~1200
)
D. AS U IUCID
:z ) S 6 7 8 9
D/II
Fig. 1. Scores shown by regions of the response surface
Most recently, the F-18 aircraft has been designed with two
controls, one for each hand. The control for the left hand has the
primary purpose of controlling thrust, and the control for the right
hand has the primary purpose of pitch and roll control.
190 W. W. WIERWILLE
The right hand control for the F-18 is similar to that de-
scribed by Woodson (1981). The control itself (Figure 3) is a two-
axis joystick for pitch and roll control of the aircraft. It con-
tains seven auxiliary control functions, three of which are per-
formed by the right thumb, another by the thumb and index finger,
another by the index finger alone, and still two others by the re-
maining fingers. Again, the types of controls used for the auxil-
iary functions vary.
A great deal of time, effort, and talent has gone into the de-
sign, development, and testing of these controls. Clearly, they are
effective and appear to work efficiently. While there may be small
revisions in the design as more experience is gained, the design
philosophy is likely to continue and to become more widespread.
Momentary
pushbutton
Momentary
pushbutton
Three-position slide
switch, momentary in Momentary
one direction, pushbutton
hold in center and
in other direction,
(Bang-bang)
"'",--B'-:::~~~"1
' \ =,"01
FRONT
(away from pilot)
Four momentary-position
center-off pushbutton
Four-position
momentary switch
(Two-axis bang bang)
Momentary
Pushbutton
Three-position toggle
switch and pushbutton
Two-axis
control stick
Two-detent
lever
switch
Momentary
pushbutton
While not all of the above are absolutely discarded, they are
to an extent subjugated to a new design philosophy.
What are the new principles that come into being along with
this new design philosophy? They appear to include the following:
CONCLUDING REMARKS
ACKNOWLEDGEMENTS
REFERENCES
Colin G. Drury
ABSTRACT
Three types of interface between the human hand and the world
of machinery are distinguished. They are handles for moving objects,
handrails for steadying the human and controls for transmitting in-
formation to a machine. For each type of human/machine interface the
shape, size and texture of the handle need to be considered. The
literature is reviewed on handles, handrails and hand controls, giv-
ing consistent recommendations on interfaces which fit the hand.
Handle position and angle are considered for two types of tasks,
manual materials handling and push/pull tasks. Both laboratory and
field studies show that handle positions on boxes should encourage a
'diagonally opposite' grip which has both horizontal and vertical
stability. Handle angle should be such as to minimize radial and
ulnar deviations of the wrist. For pushing and pulling tasks the
handle should be 900 - 1100 mm above ground level and foot obstruc-
tions should be avoided to allow free walking while pushing and
pulling.
INTRODUCTION
Where the hand meets the machines and objects of the worla of
work is an obvious, if complex, inferface. We have many words for
these hand/machine interactions: touching, holding, pushing, grasp-
ing, pulling, steadying. The machine part of the interface can be of
three types:
195
196 C.G.DRURY
....
~
198 C.G.DRURY
Ulnar/Radial
Deviation Torque
Thrust \ \
Force
cessary for the production of other forces and torques. The names of
the torques have been chosen on the basis of the action necessary at
the wrist to exert them; the actual muscles used to exert all torques
and forces will of course vary with the nature of the task.
HANDLE/CONTROL DESIGN
HANDLE SIZE
HANDLE DIAMETER
15 6.8 6
15-20 6.8- 9.1 13
20-40 9.1-18.2 19
40 18.2 25
HANDLE SHAPE
HANDLE POSITIONS
Fig. 5. Tray Designed for Automated Equipment, Not for Human Hands
204 C.G.DRURY
OVer the whole set of 2038 box handling movements, the great
majority (1837 or 90.1 %) were two handed throughout. In a further
105 movements (5.2 %) a change was made from two handed to one hand-
ed or vice versa. The remaining five percent of cases were almost
equally composed of one object handled in one hand or two objects
handled one in each hand. For this two-handed movement data there
were 81 possible pairs of hand positions but only about two-thirds
THE HAND-MACHINE INTERFACE 205
HANDLES
23 24 25
Handle positions DOD Special _ _ _ _ _ _ _ _ _ _ _-1
0
Reach Over?
Twist L or R?
38
Oft. ft. L! ft. [jl:j
3
6
2
5
1
4 > LH RH Against
Tia
Top/
=a =a
Stage Posn? Elbow Posn? ElbOlol'
9 8 7
0 0 BodeY
_a
Pre-Grasp'
- _CJ
0 _[j [j -Cj
_a _a
~ick Up
1
~ 2 3
t'lDve/Carry
lPut Down
[j
t) _a LJ
0
_15
_0 _a
_L!
_0
4 5
8
6
~djust [j _0 [j _0
n
_0 _0
7 9
DODD
77 78 79 80
1 2 3
456
789
~
Fig. 6. Manual Materials Handling Survey
206 C.G.DRURY
Subject Box
,.n3:
:::t
Z
m
Hand Position Frequency Height Weight Height Width Length Weight
Z
3/7 799 LIGHT HIGH WIDE SHORT LIGHT -I
m
::D
1/9 168 tall HEAVY LONG light ."
,.
8/8 159 TALL HEAVY LOW NARROW LONG HEAVY n
m
7/9 145 SHORT LIGHT LOW narrow SHORT LIGHT
2/8 87 TALL HEAVY high LONG
4/4 87 SHORT LIGHT LOW wide SHORT LIGHT
4/6 67 SHORT LIGHT LOW SHORT LIGHT
1/3 58 HEAVY LONG
2/2 50 heavy HIGH LONG HEAVY
6/7 31 short LIGHT narrow SHORT light
4/5 31 tall LOW LIGHT
7/8 24 TALL LOW
5/5 23 LIGHT LOW light
II.)
o-..J
"-I
0
CO
Table 3. Pattern of Significant Effects of Task Variables on Hand Positions for Move/Carry.
Upper Case Represents p < 0.001, Lower Case Represents p < 0.05
3/7 799 HIGH LOW LOW NEAR near DIFF DIFF DIFF
1/9 168 far far
8/8 159 LOW high FAR NEAR NEAR DIFF diff diff
7/9 145 long high HIGH HIGH FAR far EASY EASY EASY
2/8 87 short LOW LOW low far DIFF
4/4 87 high HIGH high FAR FAR EASY EASY EASY
4/6 77 high near far far EASY EASY EASY
1/3 58 short near
2/2 50 low low FAR near DIFF DIFF DIFF
6/7 31 high near near EASY EASY easy
4/5 31 short high EASY EASY EASY
7/8 24 far NEAR near
5/5 23 long low FAR FAR easy
n
C)
0
::D
C
::D
-<
THE HAND-MACHINE INTERFACE 209
10
that lighter boxes have 4/4 more frequently. OVerall, box width and
distance carried have the least significant effects on hand position
with all other variables showing a complex pattern of significant
effects. It appears that the frequency of hand positions is greatly
affected by subject size and box size. Table 3 shows generally that
there are more significant relationships at the start and mid stages
of the movement than at the end, suggesting that humans determine
their hand positions based on factors early in the movement rather
than factors which will be appropriate at the end.
The results showed that handle position had a very large effect
on all measures (p< 0.001) and box weight affected everything except
handle angle, heart rate and body part discomfort (p< 0.001).
But low forces do not mean low stress. Heart rate and the other
'cost to the operator' measures showed a very different pattern, for
example, that shown in Table 6. Subjective and physiological costs
were higher at both positions 3 and 8, with lower values at 6 and 9.
N
212 C.G.DRURY
The 3/7 and 1/9 hand positions, used on over 50 % of all movements,
fulfilled the conditions derived from the laboratory experiments,
providing both horizontal and vertical stability. These two diago-
nally opposite hand positions took this principle as far as pos-
sible. Hand positions were not random, as relatively few positions
accounted for the great majority of movements and hand positions
tended not to change after the pre-grasp stage until the adjust
stage. In this latter stage, hand positions on the subject's side of
the box were favored.
A low box height was associated with hand positions 8/8, 7/9,
4/4, 4/6, 4/5, 7/8 and 5/5, all in the lower half of the box. Unless
a box was tall, the subject was unlikely to benefit from having a
steadying hand near the top. Almost the same set of hand positions
were associated with narrow boxes, that is, those which are compact
fore-and-aft.
The Coury and Drury laboratory two handed holding task used
freely pivoting handles so as to measure the most natural angle of
each handle. Not surprisingly, the handle angles measured were near-
er horizontal on the bottom of the box (position 8) but became al-
most vertical along the front of the box (position 3). The actual
mean handle angles to the horizontal axis of the box were
Position 3 83°
Position 6 75°
Position 9 65°
Position 8 55°
CONCLUSIONS
ACKNOWLEDGEMENTS
REFERENCES
Harry L. Snyder
Blacksburg, Virginia/USA
INTRODUCTION
This paper will survey the current data and knowledge in the
human factors/ergonomics literature pertaining to the design and
evaluation of visual displays and visual display units (VDUs). The
discussion will be broken into several critical areas of current
concern and activity. For each such area, existing data and know-
ledge will be generally summarized, data gaps will be indicated,
current activity in research will be defined, and future problems
or requirements will be suggested. Much of the discussion will per-
tain to activities within the United States, although some techno-
logical developments from other countries will be noted. In addition,
since much of the emphasis on VDUs emanates from European countries,
these driving considerations will also be described.
219
220 H.L.SNYDER
NEW TECHNOLOGIES
Additional problems with the gas discharge panel deal with its
limited color capability. The most popular version, with a neon gas,
emits a narrow band orange color, which is totally invariant. While
experimental models have been made with several colors, the colors
are low in saturation and fairly dim in luminance. To date, these
have been considered unacceptable for any meaningful application.
Electroluminescent Displays
Ergonomics Problems
In each of the above technologies there are general problems
which exist and which have yet to be addressed adequately in the
research community. First of these is the overall size requirement
for such displays. Manufacturers have absolutely no understanding
of the combination of display size and information density which is
required for numerous applications. At the present time, there ap-
pears to be a competition to see which company can make the largest
of any given type of display, with no great concern for application,
human visual system requirements, system design requirements, or
other driving influences. Questions are often asked in an attempt
to define the total number of picture elements (pixels) needed for
a given display, without regard for its application. Similarly, it
is often aked how large a display the human visual system can "use".
Professional persons in the ergonomics field understand that these
are meaningless questions, but the enthusiasm of display designers
is not at all dampened by appropriate replies. Thus, some research
is necessary to attempt to define "rules" for display size require-
ments for various types of applications and for various display
technological types.
ERGONOMIC DATABASE FOR VISUAL DISPLAYS AND VDUs 223
COLOR DI SPLAYS
Ergonomic Issues
INTERFACE UNITS
Ergonomics Issues
The last few years have seen strong international emphasis upon
the development of standards for VDUs and VDU operators, particular-
ly under heavy usage conditions. West Germany has generated DIN
66 234 and a similar standard from the Trade Cooperative Association
(Berufsgenossenschaften). It is generally assumed that terminals
sold for commercial usage in West Germany and in much of Western
Europe must meet such standards. These standards exist because of
great concern over the visual and health problems of the worker, and
reflect the very strong union involvement in hardware design for the
workplace. Internationally, it has been estimated that approximately
15 million workers currently use VDUs, and this number is likely to
double approximately every two or three years. Because of the many
complaints of visual fatigue and other physical difficulties, it is
extremely critical that VDUs be designed to be compatible with human
visual requirements.
Ergonomics Issues
-;;;4r----r----,-----r-----,
~~
~i= 3
~~
a::...J
WW
Il.!a 2
WID
~Z
1-«
«Z~
W(f)
1
:;:!;~
~
O~----'------'------'-----.l
0.0 .05 1.0 15 2.0
INTERELEMENT SPACE/ELEMENT SIZE
(EDGE-TO-EDGE)
THREE-DIMENSIONAL DISPLAYS
to the use of the green and red glasses and putting one image on a
red phosphor and the other on a green phosphor of a traditional
color display. Other techniques to achieve 3-D capabilities in a
two-dimensional display include mechanical shutters, one for each
eye synchronized to alternate fields of the display, each field re-
presenting one of the two images necessary to achieve the third
(depth) dimension. Piezoelectric shutters are also used in place of
mechanical shutters for the same synchronized effect.
Ergonomics Issues
During the past decade, new approaches have been taken toward
the quantification and characterization of the human visual system.
This research has yielded concepts and measurements of visual system
performance which are in a form that is compatible with the analysis
and specification of visual display products, thereby permitting the
visual system and the display product to be analyzed compatibly
through linear systems analysis. The fundamental concept of interest
in the spatial domain is that of the contrast sensitivity function,
which relates threshold contrast capabilities of the visual system
to the spatial frequency (e.g., cycles per degree) of a sinusoidal
stimUlus pattern. The general result of such research is that the
visual system is essentially a bandpass filter, with the greatest
sensitivity (requiring least contrast) at approximately 3-7 cycles
per degree and requiring greater contrast at either greater or lesser
spatial frequencies. Changes in this contrast sensitivity function
have been related to such parameters as retinal illuminance, ex-
posure time, display size, observer age, grating orientation, chro-
matic grating characteristics, and the like. Further, combining the
contrast sensitivity with measures of contrast rendition (modulation)
by the display, using the same coordinates, has resulted in accept-
able measures of overall display image quality '(e.g., Biberman,
1973; Snyder, 1973; Snyder and Maddox, 1978; Task, 1979).
1.0
0.5
C~"~""
I
UMITS I I
I I
l- I
I I
I'" /
(fJ I I
<[ I
0.01 I
0::: I
l- I
Z
0
'.... ....... _-", ,;/ I
0.005 I
U I
/
"
" " ----" ... '" '"
0.001 L--'---L--'-J....L.LJ...1J_-'--'---LJ....L.JUi.1_-'---'---LLLl-U.J
0.1 0.5 I 5 10 50 100
Ergonomic Issues
SUMMARY
REFERENCES
A. van Meeteren
INTRODUCTION
235
236 A. VAN MEETEREN
.... ; . , ... .
' :"
". j.~.
. 0:1
'- :i . ..>'-:.:.'..." " , ' .;"
• -'7>- '! ~ '.
...
.;":,
'J. .. ~~~ ..
';
..
. ~
..\0
' :\.
:.. :
~.~ ,.
:
"} : . 'j' •
· .!. .....
f ·.~ .,..~-:"
, ,•
.......~:..; · . I~
.
.;..,.~ :
... ~
.....:- .
....'".
,. . .. . .
~
.\
.... <K
... .''
.\~:,
",.
..... -:' . .
"~~ ".
' , " ~F ,,\. ; I .~
.
~
.. ;;~~ :.
~. ... .:,.. ",, . ..: ;~..;'. ., ' J . :~\ ' ~
.'
:~.~. ;V , :P'i~.~ "-' " .t• .
: .;'.;:'
:...... .... ..-,.,.
.. .~ ~~~.
, ..
-
~
.
"
~"; -: .
.~! .~ . ....:. . '~~
,--..,1.", . ~;:4<0' . ....:!-.. . .
~' . ,'
·.:",r· 0
#
r'.1'.'
'; ' "",
,!f::
",'1 :
...~:.. :{~~,..
.~:~. :
."'........ .~:~.~~ .
Search
Tables
Ravenswaaij 305 xx o
FAIR Rijswijk 309 x o o
Maurik 325 x o o
Eck 229 x x x
Lienden 229 x x x
Ingen 259 x x x
Ommeren 269 xx x o
GOOD Echteld 270 x x x
Buren 299 o x x
Zoe len 329 x x x
(20 makes)
Drumpt 655 x x x
Beusichem 660 x x x
Erichem 790 x x x
Zoelmonc 279 xx x x
Deil 360 x xx x
Geldermalsen 495 xx xx x
EXCELLENT
(6 makes)
Kerkdriel 585 x xx x
Avezaath 620 xx xx xx
Wadenooijen 605 xx x x
More rules for good table design may be found in textbooks. How-
ever, if you take the functional approach, your own commson sense
will tell you how to proceed.
Illustrations
the ranges of the scales, the frame of the picture, the symbols and
the letterings, to mention a few issues, treated adequately in text-
books. Here I would like to emphasize the basic ergonomic principle,
i.e. to ask what use is to be made of the illustration? Literally an
illustration is made in order to "enlighten" something, like Fig. 2
illustrates. Graphs, for example,should not be used to present only
the data. The gain of graphical presentation is in the ordering of
the data, the quick communication of preprocessed data, in order to
facilitate "do it yourself" conclusions by the reader as mentioned
above with regard to tables. I hope the difference between the two
alternative presentations of the same data in Fig. 3 may illustrate
this point sufficiently. The original example in Fig. 3 was made
after a real graph in an otherwise excellent paper in one of our
respectable scientific journals.
Maps probably are the oldest printed visual displays for job-
aid purposes in travelling and geo-poli tics. No wonder that the
functional approach is nothing new in map design, witness the fact
that different maps were made for different purposes early in
history.
10
,
single double single double
-°e
...0
u 8 ~ 0---0
.=.
0
~
c 6
~ ~
"0.s::.
;:
III
...
QI
4
2
~ ~ -.
~ ~
0
~ 002 sec
~B to sec
A B A B A A B
~
o
10 10eo
[] '0
c 8 [] 8 c
°e [] []
°e
" 6
[]
... . 6 6:2
~III []
• !e
•
6
[]
[]
• • 4 III
~ 4 • 0
0 ...
;:
.
6 6 0
:= •
!c
0
o
! 2 o 2
QI QI
~O
°iii
o o
A B
O:go
A B A B A B
02sec 02sec 10sec 10sec 02sec 02sec 10sec 1osec
"
Fig. 3. Examples of a wrongly designed graphical presentation of
e~erimental result (below) and its improved version
(above). Different symbols refer to different subjects in
a threshold experiment. The difference between A and B in
3 of 4 conditions is better illustrated in the improved
version. Moreover, we now see immediately that all sub-
jects contribute to this difference.
Forms
Netherlands, they are bad examples more often than good. Fig. 4
shows how easily Wagenaar (personal communication) could improve
the form used by Utrecht's Electricity Company for billing the
energy delivered to domestic customers. These bills should help the
customer to monitor his expenditure and to check whether or not he
agrees. However, these forms are frequently designed to satisfy the
needs of the company. This must be kept in mind as a general issue:
who's job is made easier, the laity's job or the professional's
job?
Hard copy data presentations: more than setting soft copies hard
c
... --.. --
o
--- -,
E F I B A G
!
.
.- -
.,."'''',..~
I I
I
---
!
~I 1'1 ~tW'K-':~ -1 v.-r::_ -1. !l. o....a.-"" ~ tOl-... . .
,I
" II ,.1 1032 1 .I lZ:6()i 1 2292
~
A V M£ETEREN H
3769 DJ KAMPWG 102 SOESTERBERG 5671
04555 bC 37179!
I
A B c o E F G H
\
meter readi ng I Amountl kWh price
Amount owed VAT
Cat. last previous I used 'basic fuel total
ORIGINAL
1. Title
2. Contents
9. As 6 for 1977
12. Discussion
13. Discussion
ALTERNATIVE
4. Implementation of corrections
5. Lay-out
6. Printing.
Speeding up interactions
@) .A
....0
0-
D Is
Support of representations
more the operator is aware that there are more ships in the sector.
Moreover, he may see almost immediately that there are ships in
four different categories. VDU-screens offer a number of tricks to
call attention for data in the context of other data, as well as to
group data in the context of other data. One can apply different
light intensities, reversed contrast, underlining, flicker, or a
different colour.
that colour can be used for coding the value of a variable along a
continuous scale. It only helps to distinguish at best 7 categories.
13 4~
17 4~-
COtl1AHOO
Process control
Problem solving
However that may be, the important question now is, if VDU's with
computational powers may help us even more than pen and paper. I
have noticed at least five possible extensions:
REFERENCES
Wolfgang Dzida
St. Augustin/FRG
INTRODUCTION
During the last decade there has been much research effort in
the field of human factors of workstations with display terminals.
In Germany some of the research findings have influenced the ergo-
nomic standardization of display work stations. Strong recommen-
dations have been made, for instance, for the display contrast of
negative image displays, or for the relation between background
luminance of the display and the greater surrounding luminance of
the workstation. These human factors recommendations deal with hard-
ware characteristics of the visual display. However, it is question-
able whether an ergonomic design of the hardware-related features
will satisfy the goal of a user-friendly workstation.
257
258 w. DZIDA
Evaluation criteria can also be taken as objectives in the
field of software design. Within this context they may inspire the
system designer to think aQout prospective advantages or disadvant-
ages of a design decision. The automation of human tasks by elec-
tronically controlled systems may have some undesirable human con-
sequences, for instance, deskilling of work. That means, vocational
experience might become useless due to the fact that essential parts
of a user's. tasks and duties have been delegated to a computer. How-
ever, some rationalization effects may vanish, if the user is not
able to adapt a computer to an environmental change. The user may
become totally expert-dependent. This may lead to cost-intensive
situations because of an increasing lag between the pace of the en-
vironmental change and the limited pace of user response.
1. self-descriptiveness,
2. user control,
3. ease of learning,
4. problem adequate usability,
5. correspondence with user expectations,
6. flexibility in task handling,
7. fault tolerance.
One may argue that the above mentioned evaluation criteria are
vaguely formulated; the measurement of interface characteristics
should demand precisely defined dimensions or scales. I fully agree
with this argument. Nevertheless, I would like to illustrate some
of the difficulties which arise when the designer attempts to de-
velop a measurement method in the field of software ergonomics. Card
et al., 1980, have developed a model to measure the time it takes a
user to perform a task on an interactive computer system. User per-
formance time is simply measured by "the time it takes to do the
keystrokes" given a so-called unit task. This time is determined by
the sum of independent time components: the time for a button push,
the time for the positioning of the hand on the device, the time for
deciding which command to call, the system response time, etc.
Instrumental relationships
Causal relationships
Productive relationships
Hierarchical-sequential relationships
1. to file,
2. to alter,
3. to terminate,
4. to oversee.
Procedural relationships
no longer the user's duty. Thus, he saves much time. He can carry
out a lot of procedures when appropriate combination tools of this
kind are available.
One must take in consideration that the introduction of pro-
cedures might have some undesirable side effect. The working environ-
ment of a user is in danger of being restricted to the seemingly
"one right way", entailing lack of flexibility in task handling. It
is quite certain that the day-to-day work problems will change with
respect to goals and execution conditions. A user may be doomed to
a breakdown unless he is not able to adapt his tools quickly. If he
repeatedly has to call for an expert in order to adapt his procedures,
TOWARDS AN ERGONOMIC DESIGN OF SOFTWARE TOOLS 271
The purpose and design principle of this tool should not, how-
ever, be transferred without further ado to the design of software
tools. The objects to be manipulated by this combination tool would
not constitute a uniform productive relationship, nor would the user
recognize any hierarchical or sequential relationship between the
activities performed. The idea behind this tool is probably related
to a need of the user for easy transport. A portable computer would
have been designed according to this idea. However, for workstation
design, such a tool would cause some undesirable ergonomic side ef-
fects. The user would have to accomplish merely subsidiary activities
with results being totally disconnected. The user would never re-
cognize for what purposes the partial results are suitable. Therefore,
heterogenous productive relationships are unlikely to provide a
TOWARDS AN ERGONOMIC DESIGN OF SOFTWARE TOOLS 273
SUMMARY
Karl H. E. Kroemer
Blacksburg, Virginia/USA
275
276 K. H. E. KROEMER
In the past, much ergonomic research has been based upon ob-
servations of how person-machine interaction developed after the
system was established. Thus, ergonomic data usually consider the
reactions of people within established systems. Ergonomics is moving
away from its descriptive phase, in which much work was done in retro-
fitting and curing. Ergonomics is moving into the prescriptive phase,
which provides data and recommendations for the design of future
effective, safe and satisfying work systems.
277
278 NAME INDEX
Smallwood, R.D. 37
Snook, S.H. 214,217
Snyder, H.L. 219,223,225,228,231,
234
Spillers, W. 71,75
Squires, P.c. 147,157
Stark, L. 237,254
Stein, W. 74
Steinfeld, E. 201,217
Stewart, T.F.M. 274
Stier, G. 150,156
Stoudt, H.W. 103,113,114,143,149,
157,216
Swain, A. 73,75,80,81,92
Szostak, H. 52,55
Ulate, C. 215,217
Zangenmeister, C. 23,30
Zegeer, C.V. 87,92
SUBJECT INDEX
281
282 SUBJECT INDEX