Główna
A I I E Transactions The Control Effect of Small Weights on HandArm Movements in the Horizontal Plane
The Control Effect of Small Weights on HandArm Movements in the Horizontal Plane
Konz, Stephan, Rode, VijaykumarJak bardzo podobała Ci się ta książka?
Jaka jest jakość pobranego pliku?
Pobierz książkę, aby ocenić jej jakość
Jaka jest jakość pobranych plików?
Tom:
4
Język:
english
Czasopismo:
A I I E Transactions
DOI:
10.1080/05695557208974854
Date:
September, 1972
Plik:
PDF, 449 KB
Twoje tagi:
 Najpierw zaloguj się na swoje konto

Czy potrzebna jest pomoc? Zobacz instrukcje dotyczące wysyłania książki do Kindle
Plik zostanie dostarczony na Twój email w ciągu 15 minut.
Plik zostanie dostarczony do Twojego Kindle w ciągu 15 minut.
Zauważ: Musisz zweryfikować każdą książkę wysłaną do Kindle. Sprawdź w swojej skrzynce odbiorczej wiadomość email z potwierdzeniem od pomocy technicznej Amazon Kindle.
Zauważ: Musisz zweryfikować każdą książkę wysłaną do Kindle. Sprawdź w swojej skrzynce odbiorczej wiadomość email z potwierdzeniem od pomocy technicznej Amazon Kindle.
Powiązane listy książek
0 comments
Możesz zostawić recenzję książki i podzielić się swoimi doświadczeniami. Inni czytelnicy będą zainteresowani Twoją opinią na temat przeczytanych książek. Niezależnie od tego, czy książka ci się podoba, czy nie, jeśli powiesz im szczerze i szczegółowo, ludzie będą mogli znaleźć dla siebie nowe książki, które ich zainteresują.
1

2

This article was downloaded by: [New York University] On: 17 October 2014, At: 23:01 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 3741 Mortimer Street, London W1T 3JH, UK A I I E Transactions Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/uiie19 The Control Effect of Small Weights on HandArm Movements in the Horizontal Plane a Stephan Konz & Vijaykumar Rode a b Kansas State University , b Amstae Corpohation , Published online: 09 Jul 2007. To cite this article: Stephan Konz & Vijaykumar Rode (1972) The Control Effect of Small Weights on HandArm Movements in the Horizontal Plane, A I I E Transactions, 4:3, 228233, DOI: 10.1080/05695557208974854 To link to this article: http://dx.doi.org/10.1080/05695557208974854 PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sublicensing, systematic supply, or distribu; tion in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/termsandconditions The Control Effect of Small Weights on HandArm Movements in the Horizontal Plane  S T E P H A N KONZ Senior Member, AIIE Kansas State University V I J A Y K U M A R RODE Downloaded by [New York University] at 23:01 17 October 2014 ASSOCIATE MEMBER,AIIE AMSTAR CORPORATION Abstract: The control effect of weights of 0.12, 1.25, 2.38 a n d 3.25 pounds on speed, accuracy a n d physiological cost of handarm motions indicated: 1) Time increased 6% per pound; 2) Subject weight had n o effect; 3) Crossbody movements should be given 7% additional time. .Literature on the control effect of weight on handarm motions is quite scarce although some research seems to have been done by the Work Factor, hfethods Time Measurement (RilTM) and Basic Motion Time groups. I t is not clear whether the additional time for movements with weight recommended by the predetermined time systems are for control effects, fatigue effects or a combination. I n the Work Factor systems, additional time is given for movements which include '(work factors'' such as weight, change of direction, care, etc. The same additional time is given for a work factor whether the work factor was due to weight, change of direction, etc. A work factor increases the time for the basic movement by a percentage; the percentage varies with the number of work factors. One work factor increases the basic time by 4045y0, two by 8085% and three by 120125%. The amount of work factors for a particular weight depend on whether the operator is a man or woman. For women, no work factors are given for weights up to 1.0 pounds, one for weights between 1.1 and 3.5 pounds, two for 3.6 to 6.5 pounds, and three for 6.6 to 10 pounds. For men, no credit is given for weights up to 2.0 pounds, one work factor for weights of 2.1 to 7 pounds, two for weights of 7.1 to 13 pounds and three for weights of 13.1 to 20 pounds. As can be seen, the effect of weight on women is considered to be twice that of men. Thus a one pound weight increases time by 0% for both men and women; a two pound weight increases time by 0% for men and 40% for women; and a three pound weight increases time by 40% for men and women. Thus work factors are considered to be additive and weight is considered to have less effect for men than women. I n the MTM system, all movements with the hand empty are considered to be "reaches," and all with something in the hand are "moves." The effect of changing direction, care, etc., which was compensated for by work factors in the Work Factor System, is compensated for by categories in the reach and move tables. There seems to be no constant differential between reaches and moves either in absolute or percentage terms. At short distances, a reach from one hand to the other takes longer than a move. But, a t a distance of 20 inches, a move takes 47% more time. When reaching for an object in a fixed location is compared with moving an object to an exact location, the increase for moving is usually about 40%. It increases to 75y0 for an 18 inch move. And, when reaching for a single object whose location may vary from cycle to cycle is compared with moving an object AIIE TRANSACTIONS, Volume 4, No. 3 to an approximate location, the RITM time for moving is 5% greater than the A4TR.I time for reaching although the percentage decreases with distance. For moves which have over 2.5 pounds in one hand (10) (13), the RITM tables give additional time. This time is subdivided into a static component (for obtaining control) and a dynamic component (additional time for the travel). In the exact formula, the static 0.345 (weight, lb.). time component equals 0.475 The dynamic time increment is 1.1% per pound, irrespective of the distance moved. The Basic Motion Timestudy (BAIT) system (3) makes no distinction between reaches and moves in the basic table. There is a supplementary table for "force factors." Force factors are given for gaining control of an object, for acceleration and for deceleration. The timeperforce factor depends on the weight of the object and the distance moved. Ayoub (2) found that as weights increased from 1.5 to 3.5 lbs., maximum acceleration and maximum velocity decreased. He recommended that additional time for additional weight be added continuously rather than in intervals. The present study was concerned only with dynamic movement time; that increase in time required to move an object already under control and remaining under control a t all times. The control effects due to weight were studied but not fatigue e$ects. More detail is given in Rode (14). Downloaded by [New York University] at 23:01 17 October 2014 + Method Task: The subject alternately tapped two metal targets with a metal tipped stylus (Figure 1). The distance between the two 1inch targets was kept constant a t 16 inches. The subjects moved the stylus a t seven angles (0, 30, 60, 90, 120, 150, and 180 degrees, where zero is "1" in Figure 1) with styli of four different weights (0.12, 1.25, 2.38 and 3.25 pounds). All styli had the same external shape (one inch diameter tubing, 11 inches long with the front 1.5 inches tapered to a brass tip). The difference in weight was due to the material (wood, aluminum, steel, and steel with a lead weight). The height of the targets was one inch below the right elbow of the standing subjects; the center of the inner target was four inches from the edge of the table. The subject kept his hand and arm off the table. The targets and the areas surrounding them were connected electrically with the stylus, a battery, and a recorder so that each "hit" and "miss" on the inner and outer targets was automatically recorded. The task was performed while the subject stood on a force platform (9). The platform separates forces into their vertical, frontal, and lateral components, then records them on a strip chart recorder. A basal line for each axis was determined for individual subjects before each condition. The deviation from the basal line was traced with a planimeter and the total area (positive and negative from basal) established as an index in poundseconds, of the physiological cost exerted in each axis. The poundseconds from each of the three recorded axes were added arithmetically to determine the overall index of physiological cost. They were not added vectorially since positive and negative effort do not cancel and there is no evidence that physiological costs add vectorially. Subjects: Eight righthanded male graduate students ranging in height from 62 to 70 inches (mean 67) and weighing 104 to 160 pounds (mean 138) had forearm lengths from 17 to 19 inches (mean 18.3), and upper arm lengths from 13 to 16 inches (mean 13.9). Experimental Procedure: The seven angles and four styli made a total of 28 conditions. During the pracperformed one trial a t each tice session, each subject condition. During the experimental condition, they did two trials a t each condition. Using one stylus, they performed a t the seven angles in a different order. After the last stylus, they repeated the styli in a mirrored (reversed) order with a different randomized sequence of angles. Each subject used a different sequence of styli and different sequences of angles in order to balance learning and fatigue effects. The written instructions told the subject to "hit the targets as fast and as accurately as you can." After the 28 conditions were practiced, the experimental session began. Each condition took 12 seconds. After approximately 5 seconds to change the target angle, the next condition was recorded. A 3minute rest was given after the 14th condition, and a 5minute rest after the 28th. Subjects then repeated the conditions in reversed order with a &minute rest after 14 more conditions. ~h~ short work interval was designed to minimize fatigue effects. A Fig. 1. The experimental apparatus. Coiltact of the stylus on the inner target deflected the recorder pen 10 mm to the left. Contact on the outer deflected it 5 mm to the left. Errors gave deflections to the right of 10 and 5 mm. September 1972, AIIE TRANSACTIONS 229 Results  Table 1 : Analysis of variance for the three criteria Three criteria were used: rate of performance in bits, accuracy in percent of hits on target, and physiological cost in force platform output of poundseconds. Rate of Performance: Fitts ( 8 ) successfully adapted information theory to handarm motions with the formula : where I D = index of difficulty of the task, bits A = amplitude of move F, Source Subjects Styli Angles Sub. x Styli Sub. x Angles Styli x Angles Sub. x Styli x Angles Residual Total W = width of target in direction of move ** * Downloaded by [New York University] at 23:01 17 October 2014 ment. Assuming that a specific man's channel capacity (his ability to process information) is constant, motions of various combinations of amplitude and accuracy, but with the same ID, should take the same time. This has been confirmed by Crossman (7), Welford (16) and Annett, Golby and Kay (1) and others. Crossman and Welford proposed slight modifications of Fitt's formula but, since Fitt's is the standard formula, we used Fitt's. For this task, the amplitude of move was 16 inches and the diameter of both targets was 1 inch so I D was 5 bits. Thus if the subject moved in and out in one second, he had two movements X 5 bits/movement or 10 bits/second. If the reader prefers, he can divide all the reported bits/second values by 5 bitsjmovement to get movements/second. A mixed factor analysis of variance (Tab!e 1) showed that the effect of weight and angle was significant MTMexact a a < 0.01 < 0.05 df F, performance Accuracy Forcetime 7 3 6 21 42 18 126 4.5** 5.9** 5.7** 1.6 0.7 0.4 1.3 247 7** 65.6** 17.6** 1.9* 1.9** 0.9 0.7 99.6** 27.7** 13.1** 4.3** 6.0** 0.4 1. O 224  447 ( oc < 0.05), but the weight by angle interaction was nonsignificant. Table 2 gives the mean bits/second a t each angle and weight. A Duncan's New R/Iultiple Range Test showed that all four weights were significantly different from each other. Angles not significantly different are underlined by a common line. The subject by styli and subject by weight interactions were significant for two of the three criteria: performance and force. The following analysis of effects was significant for subjects on average, but the rank order of the effect of various weights or angles was not the same for each specific subject. As expected, the rank order of weight showed that the subjects with the least weight moved most rapidly. Figure 2 shows that the effect is slightly exponential but, if a least squares straight line is fitted, the in ,MTM  T a b l e '. F, Index of   t 8 J n VI ? 0 c Actual experiment " r. .+ 49  Femaleso M a l e s I 0 1 2 3 I I 1 I I 5 6 7 8 9 Weight (pounds) Fig. 2. The experimental data (solid triangles) do not agree with the recommendations of the predetermined time systems. The experimental data indicate time should increase 6% per pound moved to compensate for additional control requirements. 230 AIIE TRANSACTIONS, Volume 4, No. 3 Downloaded by [New York University] at 23:01 17 October 2014 Table 2: Information (bits/second) processed at various angles with various weightsa Weight, Angles, degrees pounds 30 60 0 90 120 150 180 0.12 13.2 12.4 12.7 12.4 11.8 11.2 10.6 1.25 11.8 11.5 11.3 11.3 11.1 10.5 10.2 2.38 11.2 10.8 10.5 10.4 10.5 10.1 9.5 9.8 9.7 9.2 10.4 10.1 3.25 10.5 10.0 Mean 11.7 11.3 11.1 11.0 10.8 10.4 9.9 &Valuesconnected by the same line are not significantly ( c < 0.05) different Weight, Pounds 0.12 1.25 2.38 3.25 Mean Table 3: Percent of hits on target at various angles with various weightsa Angles, degrees 150 180 60 0 30 120 94.0 91.8 93.8 92.8 92.4 91.1 95.1 94.1 93 5 93.3 92.4 98.5 95.6 95.1 92.6 93.8 93.7 94.7 93.6 95 5 92.8 93.2 93.6 93.9 95.4 94.4 93.6 93.4 93.1 92.9 90 88.3 93.2 95.6 94.3 92.9 Mean 12.0 11.1 10.4 10.0 10.9  Mean 92.0 94.3 94.4 93.9 93.7 &Valuesconnected by the same line are not significantly ( .: < 0.05) different Table 4: Physiological cost in poundseconds of force platform output at various angles with various weightsa Weight, Angles, degrees 0 30 60 90 180 120 150 Mean Pounds 0.12 1.08 1.40 1.69 1.96 2.28 2.36 2.45 1.89 1.25 1.36 1.47 1.77 2.23 2.37 2.53 2.52 2.04 2.38 1.70 1.77 2.30 2.54 2.83 2.88 2.92 2.42 3.25 2.36 3.30 2.33 2.90 2.73 2.87 3.35 3.33 Mean 1.62 1.74 2.12 2.40 2.71 2.77 2.80 2.31 &Valuesconnected by the same line are not significantly ( u < .05) different. crease in time is about 0.126 sec/pound or 6% per pound. The effect seems to start with very light weights. Do bigger people work faster? The Spearman rank correlation coefficient between body weight and mean time for all four weights was 0.50. This is not significant which suggests that for these subjects speed is not related to body weights. The incremental effect of weight might be related to the size of the arm in relation to the size of the weight. The weight of the hand and forearm is 2.3% of body weight (5) ; another estimate (6) is 2.5y0. The weight of the hand, forearm and upper arm (5) is 4.9% of body weight or 0.29 0.047 (weight, lb). Various formulae were tried to predict the reduction in time using weight alone and in combination with the weight of the hand and forearm or hand, forearm and upper arm. To our surprise, the best estimates were from the object weight alone. Considering the weight of the individual's arm did not aid in the prediction of movement time. + September 1972, AIIE TRANSACTIONS The effect of angle generally confirms the work of Briggs (4), Schmidtke and Steir (15) and Konz, Jeans, and Rathore (12) that crossbody movements are slower. The worst angle (180) was 19% slower than the best angle. In the previous study (12), maximum difference for the right hand was 18%. Accuracy: The percent of movements that hit the target is given in Table 3. An analysis of variance showed, as with the rate of performance, that the effects of weight and angle were significant and the weight by angle interaction was not significant. Accuracy with the lightest weight stylus was significantly worse than with the three heavier weights; the three heavier ones did not differ significantly among themselves. Although most angles did not differ significantly, the angles requiring a crossbody movement (150 and 180 degrees) were made more accurately (as well as more slowly). In (12) the authors showed that crossbody movements are made with less errors. Downloaded by [New York University] at 23:01 17 October 2014 Physiological cost: The mean poundseconds as recorded on the frontal, lateral and vertical axes by the force platform are given by weight and angle in Table 4. An analysis of variance showed, as with rate of performance and accuracy, that the effect of weight and angle were significant and the weight by angle interaction was not significant. The mean poundseconds increased with weight and all four weights were significantly different from each other. The poundseconds were least for movements toward the right (for these right handed movements) and increased as the movements became more crossbody. The effect of weight seems to increase exponentially but, if a least square straight line is fitted, the cost increases approximately 14% per pound of weight carried. Since time increases at about 670 per pound, there in an increase of about 8% that is not explained. Increased tremor would cause the recorded forces to deviate more from the basal values and would give greater recorded areas. An alternative explanation, compatible with Ayoub's (2) results, is that the 8% is due to greater acceleration and deceleration forces when weights are moved. The forcetime increased about 60 percent as the angle shifted from right to left: the decrease in time is 10 to 15y0. Thus it seems that 10 to 15 of the 60 (that is, about 20%) of the increase in cost is due to the decrease in speed; the remaining 90% is probably due to the increasing involvement of the upper arm and shoulder in body movements. Discussion Weight, as expected, does affect time. The control effect seems to increase time about 670 per pound of weight (0.026 sec/pound) and starts with quite light (0.12 pound) weights. For the specific conditions of this experiment, 670 per pound and 0.026 sec/pound are identical. What of other conditions? Without evidence, we suspect 6% is the better approach. Since the control effect does not seem to depend on the individual's weight, different weight allowances for men and women probably are not justified. Since we didn't test women, we can't prove that they should not have an additional allowance, but the burden of proof certainly should be on those who advocate differential allowances. Fatigue allowances may differ for men and women, but our data is for control requirements not fatigue requirements. Figure 2 shows the experimental times versus weight as well as time recommended by the RlTM exact method, the MTM card method, the Basic Motion Time recommendation, the Work Factor time for men and the Work Factor time for women. The key aspect to consider is the slope which shows how weight affects time. Schmidtke and Steir (15) pointed out some of these basic inconsistencies. Note that the intercept of the lines could vary depending on the pace of the subjects in this experiment or in the various predetermined time systems. If our subjects worked slower or faster than the pace of a specific predetermined time system, it would affect the intercept but not the slope. It was assumed, without evidence, that rate of work and the effect of weight do not interact, that it was not necessary to give a larger weight allowance to a person working a t a 120% pace than a person a t a 110% pace. The R'ITM approach seems to give inadequate importance to weight. The slope is too flat and, in the usual method, no credit is given for any weight less than 2.5 pounds. The BRIT approach also seems to give inadequate emphasis to weight, giving an increase of 0.018 sec/ pound instead of the 0.026 in our data. The smaller interval (two pounds) seems desirable. Work Factor seems to have an appropriate slope for males but its interval (five pounds) is too big. Their slope for women (about 0.052 sec/pound) is too great. Figure 2 showed that different predetermined time systems reach different conclusions concerning the effects of weight. I n evaluation of their merits, the advocates of predetermined time systems should furnish details such as number of observations, weights studied, number of replications, significance levels, etc. Obviously they cannot all be correct. Since so many people are affected by decisions made from them, it seems evidence (especially for fatigue) is worthwhile and should be published. In the meantime, a continuous adjustment for weight seems desirable (2). A practical compromise might be a step size of two pounds (at least for weights under 10 pounds). Although Ayoub (2) found males took longer than females to reach maximum accelerations and maximum velocities, to our knowledge there is no evidence that would indicate different weight allowances for men and women. If there is a subject effect, it probably is due to body weight. This experiment, however, indicated no difference for body weights from 104 to 160 pounds. The effect of angle is neglected by the predetermined time systems. Since one of their important functions is to encourage good methods work, some penalty (i.e. additional time) should be given for crossbody movements. H o ~ vmuch penalty? I n two different studies on the same type of apparatus, the first study (12) had a maximum difference with angle of 18%; this study had a maximum difference of 19y0. Since most moves are not a t the best or worst angle, typical angles of 45 and 135 degrees might be assumed. For those angles, Briggs (4) found a difference of 12Y0, Konz (11) a difference of 5.6%, and Konz, et al. (12) a difference of 6.9% for the right hand and 1.6% for the left hand. This study had a difference of 8.5%. A practical rule might be AIIE TRANSACTIONS, Volume 4, No. 3 to add 7% to the time if the reach or move is crossbody. In summary, the effects of weights of 0.12, 1.25, 2.38 and 3.25 pounds, on speed accuracy, and physiological cost of handarm motions indicated: 1) Time increased 6% per pound. 2) Subject weight had no effect. 3) Crossbody movements should be given 7% additional time. Downloaded by [New York University] at 23:01 17 October 2014 References (1) Annett, J., Gqlby, G., and Kay, H., "The Measurements of Elements in an Assembly Task ,; The Information Output of the Human Motor System, Quarterly Journal of Experimental Psychology, 10, 112, (1958). (2) Ayoub, M., "Effect of Weight and Distance Travelled on Body Member Acceleration and Velocity for Three Dimensiollal Moves," International Journal of Production Research, 5, 1, (1966). (3) Barnes, R., Motion and Time Study, 6th Editmion,John Wiley & Sons, New York, (1968). (4) Briggs, S., "A Study in the Design of Work Areas," PhD Dissertation. P r ~ r d ~ 19.53. ~e. (5) Clauser, C., McConville, J., and Young, J., "Welghl, Volume and Center of Mass of the Human Body," AMRIATR6970, WrightPatterson AFB, (1969). (6) Contini, R., Drillis, R., and Bluestein, M., "Determination of Body Segment Parameters," H u m a n Factors 5, 5 , 493504, (1963). (7) Crossman, E., "The Information Capacity of the Human Motor System in Pursuit Tracking," Quarterly Journal of Experzmantal Psychology, 12, 116, (1960). (8) Fitts. P., "The Information Capacity of the Human Motor System in Controlling the Tolerance of the Movement," J . of Experimental Psychology, 47, 381391, (1954). September 1972, AIIE TRANSACTIONS (9) Hearn, N., and,,Konz, S., "An Improved Design for a Force Platform, Ergonomics, 11, 4, 383389, (1968). (10) Karger, D., and Bayha, F., Engineered Work Measureman.f. 1ndnst)rial Press. 2nd Ed.. (1965). S., "Design of ~ o r k s t a t i o d s , "J . df Industrial Engi(11) neering, 18, 7, 41323, July 1967. (12) Konz, S., Jeans, C., and Rathore, R., "Arm Motions in the Horizontal Plane." A I Z E Transactions 2. 359370. (1969). (13) Rapheal, D., "A Study of Arm Movements Using Weights," M T M Report 108, Ann Arbor, Michigan, (1955). (14) Rode, V., "A Study of Effect of Weight and Direction on the Speed, Accuracy and Physiological Cost of One Hand Motions in the Horizontal Plane," Master's Thesis, Kansas State University, (1968). (15) Schmidtke, H., and Stier, F., "An Experimental Evaluation of the Validity of Predetermined Elemental Time Analysis Systems," Journal of Industrial Engineering, 12, 182204, (1961). (16) Welford, A,, IMeasurement of sensormotor performance," Ergonomzes, 3, 182230, July 1960. Dr. Konz is a professor of industrial engineering a t Kansas State University. His research interests include work design, personal cooling, and work environments. Dr. Konz holds an MS from the University of Iowa, a BS and MBA from the University of Michigan, and a PhD from the University of Illinois. His professional affiliations include ASQC, IEEE, AIHA, and Alpha Pi Mu. Mr. Vijaykumar Rode is an industrial engineer for the Amstar Corporation, Baltimore, Maryland where he is involved in plant industrial engineering. He holds a BS degree in mechanical engineering from Marathwada University, India, and an MS in industrial engineering from Kansas State University. Mr. Rode is a member of Phi Kappa Phi. 233