X/84/0000/ IEEE

Transcription

1 ERGONOMIC STUDIES IN COMPUTER AIDED DESIGN G.H. van der Heiden - E. Grandjean Swiss Federal Institute of Technology Department of Ergonomics and Hygiene CH-8092 Zurich, Switzerland ABSTRACT Although there is a large amount of research publications available on the ergonomics of Visual Display Terminals, only few authors applied human factors' thoughts in a CAD-environment. This paper describes the results of an ergonomic survey on interactive graphics workstations. Starting from an analysis of specific characteristics of CAD-operation, areas of improvement in workstation design can be deduced. INTRODUCTION Human factors engineering - often also called ergonomics- aims at providing a safe, healthy and comfortable working environment. Very often efficiency will also increase as a result of ergonomic effort. This paper describes the results of a field study on the ergonomic aspects of graphic VDT-workstations for interactive CAD-operation. The aim of the study was to find out if, from an ergonomic point of view, CAD-operating is similar to nongraphic VDT-operating. This information is crucial for the applicability of existing knowledge to CAD-workstations. Starting from the specific characteristics of CADoperation existing workstations were analyzed and areas of improvement in workstation design were deduced. METHODS The field study included an ergonomic survey on workstations presently being used in various CAD-applications. It consisted of three parts: - a work-sampling study to analyze the main visual and manual operating fields. - a survey of opinions of CAD operators with questionnaires. - measurements of lighting conditions and of dimensions of CAD-workstations. The technique of work-sampling (according to Haller-Wedel, [6]) was adapted to operators working interactively with CAD terminals. The aim was to analyze the direction of vision, the handling of documents, and the operating of tablet, pen and keyboard. The observation time was restricted to the time the operators actually spent at the workstation. The mean observation interval was 3 minutes. Video films were recorded to analyze in detail the dynamics of CAD operation. The work-sampling and the video recordings were carried out for three CADapplications: mechanical design, wiring board design and electrical schematics. Subjective impressions concerninc health aspects, CAD hardware and CAD software, and workstation layout were recorded by means of a questionnaire. 69 CAD-operators completed this questionnaire. It contained questions on physical and visual impairments, an evaluation of working conditions and an evaluation of information presentation. The questionnaire was also presented to a control group of 52 draftsmen engaged in similar engineering jobs without CAD. Measurements at the workplace covered such dimensions as seat and table height above floor, seat-level-to-desk distance, and workspace and legroom provided. It also contained a determination of the ambient lighting conditions (illumination levels) and of the luminances and the luminance contrasts in the visual field. The aim of these measurements was to compare objective measurement results with subjective impressions of the operators concerned and with existing ergonomic guidelines. THE CAD-TERMINAL The 38 CAD-terminals included in the study consisted of the basic workplace elements shown in figure I. These are: 21st Design Automation Conference X/84/0000/ IEEE

2 Table i: Departments studied company depart ment line of business CAD-application CAD-experience (years) I A engine construction mechanical design 2,5 II B electronics engineering printed wiring board design 2,5 C power plant engineering electrical schematic design 2,5 D railway plant engineering wiring diagram development 1,0 E CAD research software support 8,0 F power plant engineering plant & piping design 3,0 G high tension engineering wiring diagram development 1,0 III H electronics engineering printed wiring board design 3,5 IV J electronics engineering integrated circuit design 1,5 depart ment A B C D E F G H J Total Table 2: Groups studied n of n of age ± s ~ working on CAD(h/d) termi opera nals!tots % > ± i i0 31 ~ ± ± i i ± ± i experience (years) 0-i 1-3 > i i !

3 - an - a monochromatic graphic video display with a green phosphor. 27 were rasterscan displays, ii vector-storage displays. - a digitizer command tablet with a pen as a cursor control to manipulate the screen graphics. The tablet could also be utilized as a user-definable menu to provide means of rapid command input. alphanumeric keyboard, used as a data or command entry device. Of the workstations 30 also included a hardcopy-printer unit. A CADterminal table was a part of the standardequipment, and it contained CAD system electronic hardware. An extra reference table, as well as a special chair are being offered to CAD-users as an option. The number of workstations involved, as well as the number of operators examined, their age, sex, working time on CAD-terminals and the duratio~ of their experience with CAD are summarized in table 2. Table 2 reveals that 43% of the operators reported a daily work-time on CAD exceeding 4 hours; the remaining operators worked 4 or less hours per day. Furthermore it is shown that almost 3/4 of the operators had a more than one year experience with CAD-work. A large majority of the workstations (88%) was used by more than one operator alternatively. RESULTS Work sampling studies Table 3 on the next page shows the result of the work-sampling studies for the departments A, B and C, each with a different CAD-application. The operators' work-time is divided into ten activities. Pauses and other absence times are excluded. 11 i ' Fig.l: The CAD-workstation (I= graphic video display; 2= digitizer command tablet and pen; 3= alphanumeric keyboard; 4= hardcopy unit; 5= CADterminal table) Of CAD-operators involved in this study 65% also had a traditional drawing table. They were used primarely to prepare sketches for CAD-drawing and to check CADplots. DEPARTMENTS STUDIED The field study was carried out in four Swiss industrial companies. One of the companies had six independent divisions using CAD and was therefore subdivided. The nine resulting CADdepartments are characterized in table I (see next page). Table 4 shows the sum of the main activities. From table 4 it can be seen that the total time spent watching the screen lies between 48 and 68% of the working time at the CAD workstation. This means that the operators' attention was frequently directed to the screen and that the graphic display was being observed during 2,5 to 3,5 hours per operator per day on the average. It is interesting to compare this result with those obtained in conversational office tasks: here the glance at the display accounted for 30-40% of a daily VDT working time of less than 5 hours [2]. This means that CAD-operators tend to have longer screen-watching times than VDT operators in conversational office tasks. Opinions on graphic displays The frequent viewing of the screen, together with the perception of small details in the drawings imply the need of high quality displays. Of the questioned operators 20% thought that quality of the displayed images was rather poor, 12% criticized the display for being unsharp. It is necessary to differentiate between raster-scan displays and vector-storage displays as the latter had better sharpness and stability of the displayed images. Table 5 shows the operators' opinion about the legibility of the display. 222

4 Table 3: Results of the work-sampling (pwb= printed wiring board) studies I % of working time, when working interaetively on CAD Department: N of observations: A B c 1752,[ activity act no. mechanical design pwb design electrical schematics watching screen (without manipulation) operating keyboard watching screen ii not watching screen operating tablet watching screen not watching screen 5 17 i0 8 hardcopy manipulation on terminal table 6 <1 <i 1 on reference table 7 <1 <I <I document manipulation on terminal table on reference table various I00 % loo % loo % Table 4: Results of the work-sampling (pwb= printed wiring board) studies II % of working time, when working interactively on CAD Department: N of observations: A 1681 B 1670 C 1752 activity activity numbers mechanical design pwb design electrical schematics watching the screen (incl. manipulation) operating keyboard operating tablet hardcopy manipulation 6+7 <I <i document manipulation

5 Table 5: Satisfaction of 69 CAD-operators on the legibility of the graphics display raster scan displays (n of users= 53) vector storage displays (n of users= 16) I % "poor image quality" 25 % "characters unsharp" 13 % "poor resolution" totals (n of users= 69) Operating the keyboard As can be seen from tables 3 and 4, operating the keyboard accounted for 14 to 24% of the work time, and seems therefore rather infrequent in comparison to other office-terminals. The reason was that frequently used command-strings were integrated in the tablet menu and activated by the digitizing pen, instead of by the keyboard. This phenomenon was less pronounced in electrical schematics design (department C), because of the large amounts of non-standardized alphanumeric data which were input to the drawings. The use of two input mediums - tablet and keyboard- gives rise to interference problems. Since the tablet was the primary input medium, the keyboard was usually left at the side of tne tablet (also see fig. I). This arrangement induced a twisted position of trunk and arms in keyboard operation. Of the operators 41% said that they put the keyboard on top of the tablet in case of large keying quantities. The sum of the height of the tablet (70 mm middle of the tablet) plus the height of the keyboard (51 mm home-row) resulted in an unnatural position of hands and wrists while keying. Operating the tablet Operating the tablet accounted for 26 to 48% of the work time. Of 38 workstations 13 were equipped with inflexible pen-cables, which were often caught in the corners of the tablets. Of the operators 15% regarded the flexibility of the pen as bad. CAD-users adapted their tabletmenus to their own desires. Of the operators 35% felt a need for improvement of the tablet-menus they used. 26% stated that the menu contained a number of elements they never actually used Hardcopy and document manipulation Although 30 of 38 workstations were equipped with a hardcopy-printer, hardcopy useage was a negigible percentage of worktime. Of work-time 9-16% was dedicated to the manipulation of other kinds of documents, including drawings, CADmanuals, books and notes. In department A an important part of the manipulation of large engine drawings was done on the additional reference table. In both other departments (B and C) the majority of drawings was smaller (maximum size 42x30 cm-) and could therefore be handled on the terminal desk. No less than 73% of the operators reported that insufficient terminal desk work area was provided. Various activities The "various" activities accounted for the remaining 6-14% of work time. Eating, drinking coffee or tea, and talking with visitors or neighbours were the predominant activities in this category. These "breaks" could not be regarded as a relaxation period from a physiological point of view, as the operator often stayed alert for a system prompt. Moreover, 39% of the operators felt that CAD-work allows less real pauses than does comparable drawing without CAD. Lighting conditions The illumination levels, measured on the tablet, on the keyboard and on the reference table are reported in table 6. Table 6: Illumination levels on 38 CADworkplaces illumination levels on the workplace (Ix) median 90% range workplace tablet element keyboard reference table It is concluded that the mean illumination levels are rather low. This is primarely due to the fact that all departments took measures to reduce ambient light in order to improve legibility of the display and to avoid reflections. Of 9 departments 7 disconnected all fluorescent lamps and 6 224

6 kept window blinds closed permanently. However, if illumination is cut to levels under 100 lux, reading of documents is likely to suffer. Of the operators 22% thought that illumination levels on their workplaces were too low, especially for reading purposes. The point must therefore be stressed that the studied CAD-displays could hardly be used under normal office environmental lighting conditions. Special arrangements must be made, unless lighting characteristics of graphic displays improve. The luminances of the most important surfaces in the operator's visual field are reported in table 7. Table 7: Luminances at 38 CAD-workstations luminances (ca/m 2 ) median 90% range workplace screen background I0 element documents tablet (menu-elements) keyboard 3 I - 12 The figures of table 7 reveal rather low luminances which correspond to the low illumination levels. From the luminance levels the luminance contrast ratios were calculated and reported in table 8. Table 8: Contrasts of luminances in the visual field (% of 38 terminals measured) contrast documents:screen (n= 38) tablet:screen (n= 38) window:screen (n= 26) luminance contrasts <3:1 3:1-I0:I >I0:I maximum 23:1 18:1 1450:1 In order to guarantee visual comfort, it is recommended that the contrast ratios between screen and surrounding surfaces should not exceed 1:10. Table 8 shows that contrasts exceeding this threshold often appear in those workstations that face a window. Although luminance levels were generally low, 36% of the operators complained about annoying reflections on the screen which rendered reading of screen graphics more difficult. Health aspects The results of the questionnaire reveal both for the CAD-operators and for their control group of draftsmen a low incidence of physical troubles. Even pains in the back are reported by 6 and 2% respectively only. It is well known that VDT operators show a pronounced tendency for troubles in the neck-shoulder-arm-hand area [8]. It was therefore interesting to compare the results of troubles in the neck-shoulder-arm-hand area of CADoperators with those reported from office VDT-operators. This comparison is shown in figure 2 (see next page). The incidence of complaints of CAD-operators is lower than that of the office-vdt operators. We assume that the observed dynamic working methods, including frequent exercise, result in an absence of constrained postures in CADoperators. On the other hand, the CADoperators quite often reported symptoms of visual discomfort. The most frequent symptoms occurring daily are "eye fatigue" (15%), "burning eyes" (9%) and "blurring of far sight" (9%). According to a previous factor analysis of such symptoms the latter symptom could be explained by an impaired accomodation [9]. In figure 3 the results on eye impairments of CAD-operators are again compared with those of office-vdt operators and with the two control groups already shown in figure 2. It can be seen that the incidence of visual complaints of CAD- and of office-vdt operators are of the same order of magnitude, while the two control groups exhibit lower figures. In fact, relationships were found between lighting characteristics of VDTs and the incidence of visual complaints [9]. We assume that the reported eye troubles of CAD-operators are to some extent related to the frequent and intensive display watching, as well as to some insufficient lighting characteristics of the CAD-displays. 225

7 % of operators 60 _J neck shoulders right arm right hand I0- Fig. 2: Incidence of almost daily (=~) and of occasional(=~) pain, in five professional groups: l= CAD operators (n= 69) 2= data-entry terminal operators (n= 53) 3= conversational-terminal operators (n= 109) 4 = traditional office workers (n= 55) 5= draftsmen using drawing boards (n= 52) I I ~ ot operators eye-fatigue burning eyes painful eyes fligkering vlslon I0. lilj jj Mm Fig. 3: Incidence of almost daily (=~) and of occasional(=~) eye complaints, of f~ve profes- sional groups as in figure

8 DISCUSSION AND CONCLUSIONS The work-sampling studies revealed that the observed CAD-operators watched the screen for a large part of the working time at the workstation: 48-68% of worktime, meaning 2,5 to 3,5 hours per operator per day. This is rather more than the viewing time of office VDT operators in conversational jobs. Measurements and observations at the workstations showed low illumination levels, due to efforts to reduce reflections on the display; furthermore excessive luminance contrast ratios were not seldom. These conditions could very well be the main reason for the increased incidence of visual discomfort. It is concluded that CAD-user departments should pay attention to the proper design of lighting and of the arrangement of workstations in relation to windows and luminaires. The work-sampling study as well as the video film recordings demonstrated clearly a large variety of physical activities associated with frequent movements of trunk and upper limbs. The CAD-operators exhibit much less constrained postures and more dynamic behaviour than office VDT operators do. This is most probably the reason for the low incidence of physical complaints of CAD-operators in the neck-shoulder-arm area. Furthermore, the results suggest that for physical comfort an adjustable workstation would be advisable, this all the more as 88% of the workstations were occupied by more than one operator alternatively. Besides adjustability, special attention should be given to an adequate deposit area, to a comfortable horizontal legroom (60 cm front edge to back panel) and to a well designed chair with a high backrest. ACKNOWLEDGEMENT The stimulus and funding which made this original study of the ergonomics of interactive graphics workstations possible was provided by a grant from Computervision Corporation of Bedford, Massachusetts, USA. [2] [3] [4] Elias, R. and Cail, F. "Exigences visuelles et fatigue dans deux types de t~ches informatis~es." Travail Humain, 46 (1983), pp Fellmann,Th., Br~uninger, U., Gierer, R. and Grandjean, E. "An ergonomic evaluation of VDTs." Behaviour and Information Technology, ~ (1982), No. I, pp Grandjean, E. "Ergonomics related to the VDT workstation." Proceedings of the International Zurich Seminar on Digital Communication, Zurich, Switzerland, March 9-11, [5] Grandjean, E', HUnting, W., and Pidermann, M. "VDT workstation design: preferred settings and their effects." Human Factors, 2_5 (1983), No. 2, pp [6] Haller-Wedel, E. Multimomentaufnahmen in Theorie und Praxis. M~nchen, Carl Hanser Verlag, [7] HUnting, W., L~ubli, Th. and Grandjean, E. "Constrained postures of VDU operators." In: Grandjean, E. and Vigliani, E. (eds.) Ergonomic Aspects of Visual Display Terminals. London, Taylor & Francis Ltd., pp , [8] H~nting, W., L~ubli, Th. and Grandjean, E. "Postural and visual loads at VDT workplaces. I. Constrained Postures." Ergonomics, 24 (1981), No. 12, pp [9] L~ubli, Th., HUnting, W. and Grandjean, E. "Visual impairments in VDU operators related to environmental conditions." In: Grandjean, E. and Vigliani, E. (eds.) Er@onomic Aspects of Visual Display Terminals. London, Taylor & Francis Ltd., pp , [10] L~ubli, Th., HUnting, W. and Grandjean, E. "Postural and visual loads at VDT workplaces. II. Lighting conditions and visual impairments." Ergonomics, 24 (1981), No. 12, pp REFERENCES [I] Br~uninger, U., Grandjean, E., Fellmann, Th. and Gierer, R. "Lighting characteristics of VDTs." Proceedings of the International Zurich Seminar on Digital Communication, Zurich, Switzerland, March 9-11,