Printer Friendly

Evaluation of supermarket bagging using a wrist motion monitor.


Supermarkets are among the industries with the highest number of illness cases for disorders associated with repeated trauma. Supermarkets accounted for 6500 repeated trauma cases in 1994 (U.S. Department of Labor, 1996).

Bagging groceries is often an integral part of the cashier's job (although some supermarkets employ individuals whose primary job function is bagging groceries). There is only one known study that attempted to determine the musculoskeletal problems associated with the bagging of groceries. National Institute for Occupational Safety and Health (NIOSH) researchers defined bagging as a task of the cashier (NIOSH, 1991). The epidemiologic evaluation in the study did not reveal a difference in musculoskeletal disorders between cashiers who bagged groceries and those who did not. However, the ergonomic evaluation established that cashiers who did not bag groceries had 67% and 57% less repetitive movements for their right and left hands, respectively, than those who did bag groceries. The authors believed that large bagging areas contributed to the increased hand repetition of the baggers.

The study also reported that trunk flexion, shoulder flexion, and abduction are common awkward postures associated with the bagging operation. Placing plastic bags onto metal frames and opening paper bags also resulted in extreme wrist postures. When evaluating risk factors of cashiers, Harber et al. (1992) concluded that "light plastic bag" items might be the most problematic because of their high frequency of use and the type of grip used on them by baggers.

No known studies quantify the bagger's wrist movements to determine the extent of their awkward postures. Therefore, the primary objective of this study was to provide quantitative information about the wrist motions that occur during bagging tasks.


Experimental Design

The experimental design for this study was a three-factor, mixed-subject design. The within-subject factors were handle type (6 levels: six-pack, 10-cm square, 10-cm soft square, 10-cm round, 5-cm square, and 5-cm round) and pick-up location (4 levels: right front, right back, left front, and left back). The between-subject factors were site of study (2 levels: laboratory and supermarket) and participants (n = 19). The presentation orders of the object handles and object locations were randomized.

The dependent variables were maximum, minimum, mean, and range of wrist deviation; maximum velocity in each direction and mean velocity; and maximum acceleration in each direction and mean acceleration for flexion and extension, radial and ulnar deviation, and pronation and supination. The time required to bag each object was also a dependent variable.

Study Participants

The experiment was performed in the laboratory with 16 college students, and an additional 3 participants who were experienced baggers performed the study in a nearby supermarket. Among the 16 college students, half were men. All of the experienced baggers were men.


A wrist motion monitor was designed and constructed at Ohio State University (Marras & Schoenmarklin, 1991, 1993). The monitor consisted of three measurement devices that collect data in six directions of motion: flexion, extension, radial deviation, ulnar deviation, pronation, and supination. The measurement devices for radial and ulnar deviation and for flexion and extension consisted of a rotary potentiometer that is attached to two thin strips of metal. The measurement device for pronation and supination is attached to the distal and proximal ends of the forearm. The potentiometer measures the rotation of the distal end of the forearm with respect to the proximal end. The wrist motion monitor was attached to an interface that converted the analog signals to digital signals. The digital signals were then processed by experiment-specific software. Three channels were read by the data acquisition system at 25 Hz.

The bagging station at the laboratory was designed to simulate a bagging station at a nearby supermarket; therefore, dimensions mentioned here are for both the laboratory and supermarket sites. Figure 1 shows a diagram of the simulated bagging area. The bagging station consisted of the takeoff chute (96 cm high, 120 cm wide, 100 cm deep) and the bagging table (56 cm high, 67 cm wide, 35 cm deep).

The handle types are described in Table 1. All items were filled with weights, clay, or water to make them weigh 1 kg. The weight of the objects was chosen to be an average weight normally handled with one hand.


After attaching the monitor to the participant's arm, the monitor was calibrated. To calibrate each device, a voltage measurement was taken in at least five known positions for each participant in order to determine the degrees [TABULAR DATA FOR TABLE 1 OMITTED] of movement that corresponded to the voltage output. The data were converted from voltage to position in degrees using the calibration data for each participant. Linear models were good estimators of the relationship between degrees and voltage, as shown by the coefficient of determination, [r.sup.2], which ranged from .914 to 1.000.

After fitting and calibrating the wrist motion monitor on the participant, the participant began by practicing the bagging task, performing at least one trial with each of the six objects. Wrist positions were not recorded during the practice period, but data were collected during the subsequent trials. Experimental tasks were the same at both study locations. Participants were asked to move the six previously described objects from the takeoff chute into a paper bag on the bagging table. Each object was located in one of four locations during each trial. Each trial began when the experimenter positioned the object on the takeoff chute and said "O.K." The participant then picked up the object and placed it in the bag on the bagging table. The trial ended when the participant set the object down into the bag. This task was repeated for each of the six objects in each of the four locations. The participants were instructed to grasp a special portion of the object that was painted blue (except the six-pack, which was grasped by the finger-thumb handle). The objects were grasped on their right sides as seen by the participant.

Data analysis. Position data consisted of the average, maximum, minimum, and range (maximum plus minimum) for each experimental trial in the flexion-extension, radial-ulnar, and pronation-supination directions. A repeated-measures analysis of variance (ANOVA) was conducted on the wrist position data using SAS[R] (SAS Institute, Inc., Cary, NC) statistical software. Separate analyses were computed for each dependent variable for each direction of movement. Geisser-Greenhouse corrected values for p were used to correct for the bias associated with repeated measurements. A two-tailed t test was used for the planned comparison.

A finite impulse response (FIR) filter was used to filter the data before velocity and acceleration were computed. For the current application, the Hamming window was chosen. The Hamming window provides a compromise between the maximum stopband ripple and the transition bandwidth. The following is the equation for the Hamming window:

[[Omega].sub.M](n) = 0.54 - 0.46cos 2[Pi]n/M

where M is the sampling length and n = 0, 1, 2 ... M. The sampling frequency of the position data was 25 Hz; therefore, the Nyquist frequency was 12.5 Hz. Filtering the position data required two parameters: M, the sampling length, and [[Omega].sub.c], the cutoff frequency. A sampling length of 101 and a cutoff frequency of 7.5 Hz smoothed the position data and minimized the loss of acceleration data.

Velocity was computed by differentiating the filtered position data using a variation of Ridder's method as computed using Mathcad software (Mathsoft, Inc., Cambridge, MA). The average and maximum in each direction were computed for the velocity of each experimental trial. Average velocity was determined by averaging the absolute values of the directional velocities.

Acceleration was computed by differentiating the velocity data. The average and maximum in each direction were computed for the acceleration of each experimental trial. Average acceleration was determined by averaging the absolute values of the directional velocities.



There were no significant differences between the two sites when comparing the dependent variables for wrist position. There were some significant differences between the two sites when comparing wrist velocities. Maximum velocity when flexing the wrist, F(1, 17) = 4.55, p = .0478 (lab 45 [degrees] /s, supermarket 63 [degrees] /s), average velocity when flexing and extending the wrist, F(1, 17) = 4.97, p = .0396 (lab 21 [degrees] /s, supermarket 29 [degrees] /s), and average velocity when radially and ulnarly deviating the wrist, F(1,17) = 6.56, p = .0202 (lab 21 [degrees] /s, supermarket 32 [degrees] /s), were greater for supermarket participants than for laboratory participants. All other comparisons of wrist velocity, though not statistically significant, were greater for supermarket participants than for laboratory participants.

There were also significant differences between the two sites when comparing wrist acceleration. The average wrist acceleration in all three planes was statistically greater for the supermarket participants than for the laboratory participants: average flexion/extension, F(1, 17) = 12.12, p = .0029 (lab 82 [degrees] /[s.sup.2], supermarket 150 [degrees] /[s.sup.2]); average radial/ulnar, F(1,17) = 8.37, p = .0101 (lab 83 [degrees] /[s.sup.2], supermarket 145 [degrees] /[s.sup.2]); average pronation/supination, F(1, 17) = 4.52, p = .0484 (lab 231 [degrees] /[s.sup.2], supermarket 369 [degrees] /[s.sup.2]). Also, wrist acceleration was greater for supermarket participants when ulnarly moving the wrist, F(1, 17) = 4.69, p = .0448 (lab 669 [degrees] /[s.sup.2], supermarket 1199 [degrees] /[s.sup.2]).


The results show that six-pack objects required statistically more pronation and less supination than other objects. For example, average pronation for bagging was statistically greater than for other objects, F(1, 17) = 55.66, p [less than or equal to] .0001 (six-pack 35 [degrees], other objects 2 [degrees]; [ILLUSTRATION FOR FIGURE 2 OMITTED]). Average velocity in the pronation/supination direction was also statistically greater for bagging the six-pack objects than for all others, F(1, 17) = 14.10, p = .0016 (six-pack 73 [degrees] /s, other objects 59 [degrees] /s; [ILLUSTRATION FOR FIGURE 3 OMITTED]). Maximum acceleration in the supination direction was statistically greater for bagging six-pack objects than for all other objects, F(1, 17) = 33.90, p [less than or equal to] .0001 (six-pack 3770 [degrees] /[s.sup.2], other objects 1965 [degrees] /[s.sup.2]; [ILLUSTRATION FOR FIGURE 4 OMITTED]).

To a lesser extent than for six-pack objects, 10-cm objects required more severe and quicker wrist movements than did 5-cm objects. Maximum pronation was statistically greater for 10-cm objects than for 5-cm objects, F(1, 17) = 5.45, p = .0321 (10-cm 55.9 [degrees], 5-cm 41.7 [degrees]). Also, 10-cm objects required statistically less maximum extension, F(1, 17) = 4.54, p = .0480 (10-cm 28.7 [degrees], 5-cm 30.8 [degrees]), and statistically greater maximum flexion, F(1, 17) = 7.77, p = .0126 (10-cm 15.8 [degrees], 5-cm 11.7 [degrees]; [ILLUSTRATION FOR FIGURE 2 OMITTED]). Average velocity in the radial/ulnar direction was statistically greater for 10-cm objects than for 5-cm objects, F(1, 17) = 5.59, p = .0302 (10-cm 25 [degrees] /s, 5-cm 22 [degrees] /s). Average velocity in the pronation/supination direction was also statistically greater for 10-cm objects than for 5-cm objects, F(1, 17) = 4.59, p = .0469 (10-cm 61 [degrees] /s, 5-cm 55 [degrees] /s; [ILLUSTRATION FOR FIGURE 3 OMITTED]).


A comparison of wrist positions by location is shown in Figure 5. A comparison of wrist velocities by location showed that picking up objects from the right locations resulted in greater average wrist velocities when flexing and extending the wrist, F(1, 17) = 13.13, p = .0021 (right 25 [degrees] /s, left 20 [degrees] /s), and greater maximum velocity when extending the wrist, F(1, 17) = 9.66, p = .0064 (right 61 [degrees] /s, left 52 [degrees] /s), than did picking objects up from the left locations. Other comparisons of wrist velocity by location found no significant differences. There were no significant differences between wrist accelerations when comparing the different locations.


When comparing the time to bag each object, site was significant, F(1, 17) = 5.79 p = .0278. The average time for the task to be performed by laboratory participants was 4.8 s, compared with 3.2 s for supermarket participants. Object and location were also significant. The planned comparison showed that bagging time for six-pack objects (5.0 s) was statistically greater than for all other objects (4.4 s), F(1, 17) = 20.46, p = .0003, bagging time for 10-cm objects (4.6 s) was statistically greater than for 5-cm objects (4.2 s), F(1, 17) = 14.28, p = .0015, and bagging time for back locations (4.8 s) was statistically greater than for front locations (4.3 s), F(1, 17) = 10.15, p = .0054.


Six-Pack versus All Other Objects

The greatest differences between six-pack objects and other objects were found when comparing pronation and supination. The six-pack object required more extreme pronation [ILLUSTRATION FOR FIGURE 6A OMITTED], greater wrist velocity for pronation and supination, and greater acceleration for pronation and supination than did all of the other objects. Velocity and acceleration in the supination direction accounted for the largest differences; this fast supination movement was probably in response to the large deviation in the pronation direction required to pick up a six-pack object. The finger-thumb handle is the only object that had a top coupling; the other objects were picked up from the side. Extreme pronations, together with the higher velocities and accelerations required to pick up a six-pack, may be responsible for any cases of pronator syndrome that can be caused by repeated pronation coupled with force.

Object Size, Shape, and Pliability

The 10-cm objects as compared with the 5-cm objects required, for most comparisons, larger radial/ulnar and pronation/supination wrist positions [ILLUSTRATION FOR FIGURES 6B AND 6C OMITTED] and velocities. More extreme positions and greater velocities may be more likely to cause cumulative trauma disorders such as carpal tunnel syndrome, tendinitis, and tenosynovitis. For most comparisons there were no statistical differences between soft and solid objects or round and square objects.


The right and left locations were significantly different; the differences existed only in the flexion and extension wrist movements. The differences in flexion and extension for right and left locations may have been affected by the fact that all participants were right-handed and, therefore, were closer to the right locations. The participants moved their wrists over a larger range when picking up an object from the right side during the same amount of time; therefore, the velocity and acceleration for this task increased. More extreme movements, movements over a larger range, greater velocities, and accelerations as found for right locations may exacerbate conditions caused by repetition and force.

Many of the pronation and supination variables were statistically greater for back locations than for front locations. Front locations required greater pronation, possibly because they were closer to the participant. Back locations required a larger range of pronation and supination. Therefore, either a greater velocity or a longer time was needed to complete the task; in this instance, both the velocity and time were greater.

Close locations required either more extreme wrist flexion or more extreme wrist pronation. Therefore, the ideal location for minimizing wrist deviations may be farther away than the front location used in this study. However, extreme positions might lead to repeated bending, which may be problematic because it may cause greater moment arms for grocery items positioned farther away.

Comparison with Previous Research

Man:as and Schoenmarklin (1991, 1995) used the same wrist motion monitor to quantify wrist position, velocity, and acceleration of highly repetitive, hand-intensive industrial tasks. Their results for position data are similar to the results obtained in this study. However, Marras and Schoenmarklin found higher maximum velocities among industrial workers as compared with the supermarket baggers in this study. The average velocity in the flexion and extension direction was 22.6 [degrees] /s in this study. Marras and Schoenmarklin found an average of 28.7 [degrees] /s for industrial workers who had a low risk of cumulative trauma disorders (CTDs) and 42.2 [degrees] /s for those who had a high risk of CTDs. Acceleration further differentiated the two studies. The average acceleration in the flexion and extension direction was 94 [degrees] /[s.sup.2] for this study, 494 [degrees] /[s.sup.2] for low-risk industrial workers, and 824 [degrees] /[s.sup.2] for high-risk industrial workers.

Higher velocities and accelerations can be accounted for by comparing wrist movement for an 8-h shift. Marras and Schoenmarklin (1991) found that participants in their low-risk category performed an average of 24 738 wrist movements during an 8-h shift, and those in their high-risk category performed an average of 26 132 wrist movements during an 8-h shift. For comparison, supermarket cashiers had an average of 11 452 wrist movements during an 8-h shift (NIOSH, 1990). Additionally, Marras and Schoenmarklin sampled workers during 10-s periods of movement, whereas idle time between or within a cycle was not monitored. Therefore, the higher velocities and accelerations required for the manufacturing workers in the Marras and Schoenmarklin study than for the supermarket baggers in the present study may be attributable to the inherently higher repetitive demands for assembly line jobs.

Marras, Marklin, Greenspan, and Lehman (1995) also used the wrist motion monitor on participants performing scanning operations to compare types of scanners and checkstands. They reported only acceleration data. Average acceleration data in the flexion/extension direction was 200-700 [degrees] /[s.sup.2] in their study of scanning. Once again, workers scanning groceries require much higher acceleration because of the scanning task. There were no comparisons among types of grocery item scanned.


The more experienced participants at the supermarket site picked up the objects significantly faster than did the inexperienced participants at the laboratory. The experienced baggers also had significantly greater wrist velocities and accelerations. Therefore, experience of the participant did make a difference in the speed of wrist movements. Position data was not affected by experience level.

Objects with finger-thumb handle couplings (six-pack) required extreme pronation (the maximum average was 120 [degrees] of pronation) and high wrist velocities in the pronation and supination directions. These extreme motions, if combined with forceful gripping, may increase the likelihood of a bagger developing pronator syndrome or DeQuervain's syndrome.

The 10-cm objects required a larger range of movement and a greater velocity, which may increase the likelihood of baggers developing DeQuervain's syndrome or repetitive motion disorders such as carpal tunnel syndrome, tendinitis, and tenosynovitis.

Locations that are too close to the worker may require the worker to use more extreme wrist positions. Conversely, locations that are too far from the bagger may not require extreme wrist positions but may require extreme torso flexion and result in injuries to the worker's back.

Future Research

This research included different handle-coupling types and different bagging locations, but the weight of each object was held constant. Future research should include a similar study to see if the weight of the object affects the wrist position of the bagger. Additionally, a lumbar motion monitor should be attached to the participant's back to determine whether different objects or locations result in extreme torso positions.

More research should be done to determine if the product designer should provide a side-handle coupling as opposed to a finger-thumb handle coupling (six-pack) to reduce the amount of pronation required to pick up the object. Large objects (10 cm or more) could have a handle for the bagger to grasp. Bagging workstations should be studied to determine optimum distance so that grocery items are not too close to or too far from the worker.


For applications in which the participant moves very fast, a higher sampling speed is recommended. A higher sampling rate would ensure that no extreme positions were missed. The sample size was small for the experienced baggers, and momentum was disrupted for all participants because items were presented only one at a time.


This study was based on a master's thesis conducted by the first author at Virginia Polytechnic Institute and State University. The first author was supported by the National Institute for Occupational Safety and Health's Long-Term Training program. The authors would also like to thank J. C. Woldstad, H. L. Snyder, J. H. Jones, W. S. Marras, D. S. Watkins, and E. Wells.


Harber, P., Bloswick, D., Pena, L., Beck, J., Lee, J., & Baker, D. (1992). The ergonomic challenge of repetitive motion with varying ergonomic stresses. Journal of Occupational Medicine, 34, 518-528.

Marras, W. S., Marklin, R. W., Greenspan. G. J., & Lehman, K. R. (1995). Quantification of wrist motions during scanning. Human Factors, 57, 412-425.

Marras, W. S., & Schoenmarklin, R. W. (1991). Quantification of wrist motion in highly repetitive. hand-intensive industrial jobs (Grant Nos. 1 RO1 OH2521-01 and 02). Columbus: Ohio State University. Biodynamics Laboratory.

Marras. W. S., & Schoenmarklin, R. W. (1995). Wrist motions in industry. Ergonomics 36, 341-351.

National Institute for Occupational Safety and Health (NIOSH). (1990). Hazard evaluation and technical assistance report: Kroger Company, Oxford, Ohio (NIOSH Report No. HEAT 88-345-2031). Cincinnati. OH: Author.

National Institute for Occupational Safety and Health (NIOSH). (1991). Hazard evaluation and technical assistance report: Shoprite Supermarkets, New Jersey-New York (NIOSH Report No. HEAT 88-344-2092). Cincinnati, OH: Author.

U.S. Department of Labor. Bureau of Labor Statistics. (1996). Characteristics of injuries and illnesses resulting in absences from work, 1994 (USDOL USDL-96-163). Washington, DC: U.S. Department of Labor, Bureau of Labor Statistics.

Cheryl Fairfield Estill is an ergonomist with the National Institute for Occupational Safety and Health and is a Lieutenant Commander in the U.S. Public Health Service. She received an M.S. in industrial and systems engineering from Virginia Polytechnic Institute and State University in 1994.

Karl H. E. Kroemer is professor of Industrial and Systems Engineering and directs the Industrial Ergonomics Laboratory at Virginia Polytechnic Institute and State University. He received the Doctor of Engineering degree in mechanical engineering in 1965 from the Technical University of Hannover, Germany. rkload assess
COPYRIGHT 1998 Human Factors and Ergonomics Society
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1998 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Estill, Cheryl Fairfield; Kroemer, K.H.E.
Publication:Human Factors
Date:Dec 1, 1998
Previous Article:Measures and interpretations of vigilance performance: evidence against the detection criterion.
Next Article:Three-dimensional workspace for industrial workstations.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters