Journal Search Engine
Search Advanced Search Adode Reader(link)
Download PDF Export Citaion korean bibliography PMC previewer
ISSN : 1598-7248 (Print)
ISSN : 2234-6473 (Online)
Industrial Engineering & Management Systems Vol.18 No.1 pp.78-88
DOI : https://doi.org/10.7232/iems.2019.18.1.078

Effect of Grasp Height and Inspection Posture on Upper Limb Load during Single-handed Visual Inspection of Objects

Takuya Hida*, Yuka Yoshioka, Akihiko Seo
Department of Industrial & Systems Engineering, Aoyama Gakuin University, Japan
Faculty of System Design, Tokyo Metropolitan University, Japan
Corresponding Author, E-mail: hida@ise.aoyama.ac.jp
September 28, 2017 March 5, 2018 January 7, 2019

ABSTRACT


Automated systems based on image processing have been introduced into the visual inspection process. However, owing to persisting technical and cost issues, human visual inspection still plays a major role in industrial inspections. When workers inspect small items (e.g., digital camera lenses or containers of liquid cosmetics) with one hand, which leads to an awkward posture, they experience upper-limb loading because of high loads on the upper extremities. This upper-limb loading may damage their hands, arms, and shoulders. There are a very limited number of studies that investigate the effects of upper-limb loading on humans during visual inspections; therefore, we conducted experiments on subjects where they used one hand to handle small objects for visual inspections. Using electromyography, joint angle measurements, and subjective evaluations, we investigated the effects of the grasp height and the subject’s inspection posture on the load applied to the upper limb. The results showed that the effects of upper-limb loading varied depending on the inspection posture because of variations in the locations where the loads acted on the body. Therefore, when evaluating upper-limb loading during such tasks, it was necessary to consider not only the muscle load but also the posture, the posture duration, and the subjective metrics.



초록


    1. INTRODUCTION

    Visual inspection is a method of assessing the quality of industrial products (Cho et al., 2005) in which an inspector visually evaluates the surface characteristics of an object. Some characteristics can be defined and detected as a physical quantity, and others can only be evaluated by humans. Thus, there still exists a major need for human visual inspection. Therefore, various relevant studies have been conducted to improve inspection accuracy and inspection efficiency (Drury, 1973;Megaw and Richardson, 1979;Kleiner and Drury, 1993;Gramopadhye et al., 1998;Melloy et al., 2000;Garrett et al., 2000;Nalanagula et al., 2006;Baudet et al., 2013;Schneider et al., 2013;See, 2015;Nakajima et al., 2018). However, these studies have shown little consideration of physical workload.

    In a conventional visual inspection, the inspection is carried out with the object held in place by both hands of the inspector, with very little movement of the object. In this context, the physical workload is nominal, and most ergonomics studies therefore have aimed to reduce the inspectors’ mental workload and visual fatigue. In particular, visual fatigue and headache have been shown to be reduced by ergonomic improvements in the context of seated visual inspection using a magnifying glass: a decrease in magnifying glass usage, a reduction in the glare of the inspection template, etc. (Yeow and Sen, 2004). Other studies have clarified that the cognitive tasks cause a mental workload (Lee and Chan, 2009) and visual fatigue (Jebaraj et al., 1999;Lin et al., 2010). In addition, it is revealed that the physical workload arises from the prolonged visual fixation on objects and maintaining a working posture (Liao and Drury, 2000).

    However, in recent years, another style of inspection has emerged, requiring the grasping and handling of parts or products by the inspector. The associated repetitive motions, prolonged upper limb elevation, and awkward postures are suspected to cause more upper limb disorders than traditional styles of inspection. Therefore, Hida et al. (2012, 2013a, 2013b, 2015) studied the upper limb load during inspection entailing the handling of lightweight and large-sized objects, such as plastic parts for car bodies, using both hands. The results clarified the effects of various in-spection conditions (e.g., object size, inspection speed, and grasp position) on the muscle load in the upper limbs. In contrast, in this study, we focused on the visual inspection of relatively small objects such as camera lenses and containers of liquid cosmetics. In these inspections, owing to the small weights of the objects, this type of work does not overload the musculoskeletal system of the inspector. However, it is still considered to cause large workloads on the upper limb because it involves maintaining an awkward posture and elevating the upper limb forward while handling objects using one hand.

    Therefore, in this study, we conducted an experiment in which the subjects were asked to assume the position of visually inspecting small objects while handling them using one hand, and clarified the characteristics of workload by such task. Concretely speaking, the effect of the grasp height and inspection posture at which the object is grasped on the upper limb load during the inspection tasks was investigated using electromyography, joint angle measurements, and subjective evaluations.

    2. METHODOLOGY

    2.1 Subjects

    The subjects were six healthy male and six healthy female university students that do not experience pain in their upper limbs. The average (± standard deviation) age, height, and body mass were 22.8 ± 1.0 years, 22.2 ± 0.8 years, 173.4 ± 11.4cm, 159.9 ± 10.7cm, and 64.5 ± 10.7kg, 47.2 ± 6.5kg, respectively. All subjects were right-handed, and the experiment was conducted on the right upper limb.

    2.2 Experimental Conditions

    In visual inspection, it is necessary to visually detect a potential variety of defects. However, different defect types demand different positional relationships between the inspection object and any assistive lighting. For example, in the case of scratches on an uneven surface, detection is facilitated by lateral lighting. However, in the case of printing failure on a planar surface, detection is facilitated by frontal lighting. Since a small inspection object can be easily moved, the variety of required lighting conditions is typically accommodated by moving the product up and down or tilting it during the inspection. However, no study has yet determined objective best practices for inspecting such an object. The present study adopted three grasp heights (as defined in Figure 1) and six inspection postures (as defined in Figure 2) as experimental factors, all using the same inspection object.

    2.3 Experimental Procedure

    During the experiment, the subjects grasped a cube-shaped object using the right hand while seated. The object had a side length of 60 mm and a weight of 8 g, and it had a 4 mm-diameter, 20 mm-deep hole. One of five different symbols (star, circle, triangle, square, or cross) was placed at the bottom of the hole such that it could only be viewed from a limited range of angles (see Figure 3). The subjects were asked to inspect the object visually and state which symbol was in the object while grasping the object at three different heights with six different inspection postures. Upon viewing the object, the subject maintained the work posture for 5 s. Three operations were performed under each condition, and a total of 54 inspections were conducted. These conditions were randomized to eliminate the effect of the order of execution. Because easily identifiable symbols were used and very little misidentification occurred, the correctness of the shape identified was not considered.

    In addition, light-emitting diodes were installed on the left and right rear of the subject such that an illuminance of at least 1000 lx could be maintained at any grasping position. This illuminance was set by taking into consideration the illuminance at the time of performing an inspection at a chemical factory (which requires precise visual inspection) based on the factory illuminance standard in the Japanese Industrial Standards.

    2.4. Data Acquisition and Analysis

    In this study, surface electromyography (EMG) was used to evaluate the muscle loads in the upper limb due to the motion of the upper limb. The percentage of the maximum voluntary contraction (%MVC) was calculated as the ratio between the muscle activity during the task and the MVC. The joint angle was measured to evaluate the joint load in the upper limb while maintaining a posture. In addition to these quantitative evaluations, subjective evaluations were conducted after each experimental condition was tested. An analysis of variance was applied to the following variables: subject, grasp height, and inspection posture. In addition, Bonferroni’s test was used as a follow-up test of the main effect. The significance level was set at 5% for both tests.

    2.4.1 Electromyography

    EMG was performed using an active electrode (SX230-1000; Biometrics Ltd.). The signal was amplified 1,000 times, and the resulting raw waveform was recorded at a sampling frequency of 500Hz. The root-mean-square was calculated after removing the body motion noise using a 10Hz high-pass filter, and the resulting signal was smoothed by a 2Hz low-pass filter.

    In this experiment’s target work, workloads likely differed not only across conditions but also across ana-tomical loci: upper-limb elevation, wrist flexion and extension, and object grasp. However, the multitude of muscles related to these motions makes measuring all EMGs difficult. Furthermore, measuring the MVC for each of many muscles can become a burden on the subject. Therefore, the number of measurement loci was set to eight. EMG signals were measured in the following loci on the right side of the body: the upper part of the trapezius muscle, the clavicular part of the pectoralis major muscle, the middle part of the deltoid muscle, the biceps brachii muscle, the round pronator muscle, the flexor carpi ulnaris muscle, the extensor carpi radialis longus muscle, and the thenar muscle.

    The upper part of the trapezius muscle elevates the scapula, the clavicular part of the pectoralis major muscle performs horizontal flexion and adduction of the shoulder joint, the middle part of the deltoid muscle performs abduction of the shoulder joint, the biceps brachii muscle performs flexion of the elbow joint, the round pronator muscle performs pronation of the forearm, the flexor carpi ulnaris muscle performs flexion and adduction of the wrist joint, the extensor carpi radialis longus muscle performs extension and abduction of the wrist joint, and the thenar muscle performs abduction, adduction, flexion, and opposition of the thumb (Criswell, 2010). The signals collected from these EMG sites were used to evaluate the upper limb load while executing the tasks.

    The MVC of each muscle was measured as follows (Perotto, 1994;Criswell, 2010). The subject was asked to maintain isometric contraction while resistance was applied against the respective action of each muscle. To measure the MVC of the upper part of the trapezius muscle, the subject elevated the scapula while the arm was hanging down. To measure the MVC of the clavicu-lar part of the pectoralis major muscle, the subject per-formed horizontal adduction of the arm with the upper arm, and the forearm was flexed at 90 degrees. To measure the MVC of the middle part of the deltoid muscle, the subject performed abduction of the arm while the arm was slightly abducted. To measure the MVC of the biceps brachii muscle, the subject flexed the forearm from an angle of 90 degrees while in a supine position. To measure the MVC of the round pronator muscle, the subject held a fixed vertical bar and pronated the forearm from the neutral position, with the forearm flexed at an angle of 90 degrees. To measure the MVC of the flexor carpi ulnaris muscle, the subject performed adduction of the wrist, with the forearm in a neutral position and flexed at an angle of 90 degrees. To measure the MVC of the extensor carpi radialis longus muscle, the subject performed radial deviation of the wrist, with the forearm in the neutral position and flexed at an angle of 90 degrees. To measure the MVC of the thenar muscle, the subject flexed the thumb. The measured EMG was then used to calculate the %MVC for each muscle under each experimental condition. Then, the muscle load was evaluated as the average %MVC value from all muscles.

    2.4.2 Joint Angle

    A 3D sensor (9-Degrees-of-Freedom Razor IMU, SparkFun Electronics) was used to measure the joint angles. This sensor can measure triaxial terrestrial magnetism, triaxial acceleration, and triaxial angular velocity. Sensors were attached to the subject’s upper arm, the forearm, and the back of the hand. From these three sensors, the three-dimensional angle of each segment was estimated. To estimate the three-dimensional angle, the elevation angle and roll angle were calculated from the change in gravitational acceleration using the acceleration sensor. At the same time, the azimuth angle was calculated from the change in geomagnetism using the geomagnetic sensor. In addition, the gyro sensor was used to correct the estimated value in case an acceleration other than gravitational acceleration was recorded. The sampling rate was 30Hz. To minimize sensor drift, the sensor was recalibrated for each trial.

    In the evaluation of the joint angle, a value for the cumulative displacement of the joint angle was defined and used rather than using the measurements at individual joints. The neutral position of each degree of freedom (DOF) was defined as follows. The neutral position is the fundamental standing position when the joint angle reaches 0°, except for elbow joint flexion-extension, for which the neutral position is the position of the elbow joint flexed 90° (McAtamney and Corlett, 1993). First, for each joint, the difference between the joint angles at a neutral position and in a working posture was calculated during the 5-s maintenance of the work posture. In joints that have more than one degree of freedom (DOF), the difference between the neutral positions of each DOF and of the joint was calculated. Then, the cumulative displacement is defined as the sum of the absolute values of these differences. The reason for using the cumulative displacement of the joint angle is that experimental conditions differed for each joint whose angular measure increased. That is, when the cumulative joint angle displacement was large, then it was far from the neutral position; this made such an upper-limb posture difficult to assume and maintain, so workload on the joint was high. Measuring in terms of cumulative displacement allowed for an effective estimation of the difficulty of assuming and maintaining a given posture.

    The DOFs for each joint were defined as follows. The shoulder joint has three DOFs (abduction-adduction, flexion-extension, and internal-external rotation). The elbow joint has one DOF (flexion-extension). The forearm has one DOF (pronation-supination). Finally, the wrist joint has two DOFs (radial-ulnar flexion and palmar-dorsal flexion).

    2.4.3 Subjective Evaluation

    A subjective evaluation was conducted via a ques-tionnaire that the subjects filled out after each experimental condition was tested. The subjective physical load was evaluated on the upper limb overall and in four body areas: neck, shoulder, elbow, and wrist. The difficulty of maintaining the posture was also evaluated. The rating scale ranged from 1 (feeling no physical load/difficulty at all) to 5 (feeling a significant physical load/difficulty).

    3. RESULTS AND DISCUSSION

    3.1 Electromyography

    Table 1 and Figures 4-11 show the mean and standard deviation of the %MVC for each muscle. The asterisks in the graphs denote significant difference of 5%. The results indicate that, on average, the %MVC values of the trapezius muscle, deltoid muscle, and thenar muscle exceed 10%, indicating high muscle load, and the %MVC values of the round pronator muscle and flexor carpi ulnaris muscle are less than 5% on average, representing low muscle load. Fig. 5, 6, 7, 8, 9, 10

    In a previous study, Hida and Seo (2012) characterized the upper limb load during visual inspections of lightweight and large-sized objects. The results show that the muscle load is generally less than 10% MVC, exceeding 10% MVC only when the tilt angle of the inspection object is large. In other words, the muscle load during the inspection task in this study is relatively large. In addition, Björkstén and Jonsson (1977) have quantified the endurance limit as 8% MVC and recommended 5% MVC as a limit of the upper limb during sustained static contractions in the general population. This implies that the muscle load associated with the inspection task considered in this study is large.

    As a result of the statistical analysis, the significant main effect was observed except the round pronator muscle and the thenar muscle with respect to the grasp height, and the significant main effect was observed in all muscles with respect to the inspection posture. In addition, the significant interaction effect was observed except the round pronator muscle. Since the significant difference was observed in the interaction effect, the simple main effect test (Bonferroni’s test) was carried out. As a result, a characteristic tendency different from other inspection posture was observed in the inspection posture C. Specifically, in the upper part of the trapezius muscle, the middle part of the deltoid muscle, the flexor carpi ulnaris muscle, and the extensor carpi radialis longus muscle, the influence on the muscle load due to the change of the grasp height was smaller than in other inspection postures, and the muscle load was great at any grasp height. In addition, in the inspection posture C in the biceps brachii muscle, the influence on the muscle load due to the change of the grasp height was great, and the muscle load was maximum when the grasp height was in the downward from the eye level. In this way, the specific tendency was observed by combination of inspection posture C and grasp height. For the following parts, we discussed the main effect.

    3.1.1 Grasp Height

    The grasp height was found to influence the muscle load: the higher the grasp height, the greater the muscle load at the upper part of the trapezius muscle, the middle part of the deltoid muscle, the flexor carpi ulnaris muscle, and the extensor carpi radialis longus muscle. At higher grasp heights, the subjects needed to lift their shoulders in order to extend the upper limb farther. For this reason, the muscle load on the upper part of the trapezius muscle, which elevates the scapula, was greater. Additionally, when the scapula is elevated and rotated when the upper limb is raised, the upper limb becomes closer to the abduction position, as can be seen from the scapula. Therefore, as the grasp height increases, the muscle load of the middle part of the deltoid muscle, which performs abduction of the shoulder joint, increases. Additionally, the wrist must be held in an awkward position in order to rotate the object to be inspected in order to keep the face of the object perpendicular to the line of sight. As a result, the muscle loads of the flexor carpi ulnaris muscle and the extensor carpi radialis longus muscle, which are involved in wrist motion, are greater. The higher the grasp height is, the lower the muscle loads at the clavicular part of the pectoralis major muscle and the biceps brachii muscle become. At higher grasp heights, the muscle load of the clavicular part of the pectoralis major muscle, which acts in the horizontal adduction of the shoulder joint, is lower because the shoulder joint is abducted. Moreover, the lever arm of the moment applied to the elbow joint becomes longer when the grasp height is below eye level as the forearm is almost horizontal, thus increasing the muscle load of the biceps brachii muscle. The loads on the round pronator muscle and the thenar muscle were not affected by the grasp height.

    3.1.2 Inspection Posture

    The %MVC of the upper part of the trapezius muscle was greatest with inspection posture C. This is because the position of the wrist is higher in this grasp than in other inspection postures, making it necessary to raise the shoulder. The %MVC of the middle part of the deltoid muscle with inspection postures D and F were lower than those with the other inspection postures. This is because it is necessary to adduct the shoulder and rotate it inward for these postures. The %MVC of the clavicular part of the pectoralis major muscle was greater in inspection postures B, D, and F than with the other inspection postures. This was presumed to be because the clavicular part of the pectoralis major muscle performs horizontal flexion of the shoulder joint and these inspection postures require the subject to hold the forearm parallel to the midline. The %MVC of the biceps brachii muscle was greatest in inspection postures C, D, and E. This is because, in these postures, the forearm assumes a supination position such that the brachii muscle of the biceps can perform elbow flexion with ease. Although the forearm assumes a supination position in inspection posture B, the elbow joint moment is small, and the force required from the biceps brachii muscle is small because the forearm position is nearly vertical. The %MVC of the flexor carpi ulnaris muscle was greatest with inspection postures C and E. These inspection postures require the subject to keep their line of sight perpendicular to the surface being inspected by ulnar and palmar flexion in the wrist joint. These movements are performed by the flexor carpi ulnaris muscle, resulting in the relatively high observed load. The %MVC of the extensor carpi radialis longus muscle was greatest in inspection posture F. Inspection posture F requires the subject to maintain a line of sight perpendicular to the surface being inspected by radial and dorsal flexion at the wrist joint. These movements are performed by the extensor carpi radialis longus muscle, resulting in the relatively high observed load. The %MVC of the thenar muscle was lower with inspection postures C and D than with other inspection postures. It was assumed that this is because it is difficult to exert the grasping force because the hand is in a posture close to the resting position. The %MVC of the round pronator muscle was greatest with inspection posture B. However, since the %MVC remained below approximately 6%, it can be said that the participation of the round pronator muscle was small in the tasks considered in this study.

    By comparing the influence on the muscle load due to the difference in the inspection posture for each muscle, the difference between the inspection posture condition that was the minimum muscle load and the inspection posture condition that was the maximum muscle load is as follows: the upper part of the trapezius muscle is 1.8 times (B and C), the clavicular part of the pectoralis major muscle is 3.3 times (A and F), the middle part of the deltoid muscle is 1.7 times (D and E), the biceps brachii muscle is 3.3 times (F and D), the round pronator muscle is 3.9 times (D and B), the flexor carpi ulnaris muscle is 2.5 times (F and C), the extensor carpi extension muscle is 6.5 times (D and F), and the thenar muscle is 4.4 times (D and B). In other words, it can be said that the effect by the difference of the wrist posture is greater in the forearm muscle group than in the shoulder muscles.

    3.1.3 Recommended Conditions

    On the basis of these findings, a grasp height below eye level is recommended. This is because the muscle load tends to be reduced in the muscles involving shoulder joint motion. In addition, it was found that the muscles that experience the greatest load differ based on the inspection posture. Therefore, these values must be evaluated synthetically rather than individually. To achieve this, the square sum of all of the %MVC values was calculated for grasping height and used an evaluation index for the integral workload on the upper limb muscles (see Figure 12). The results of these calculations showed that inspection posture D had the lowest square sum. This indicates that inspection posture D and a grasp height below eye level result in the least workload on the muscles; therefore, this is the recommended grasp position. Figure 13, 14, 15, 16, 17, 18

    3.2 Joint Angle

    Table 2 and Figures 13-19 show the mean and standard deviation of the joint displacement for each DOF, and Figure 20 shows the mean of the cumulative joint displacement for each condition. The asterisks in the graphs denote significant difference of 5%. Figure 19

    As a result of statistical analysis, the main effect which is significant difference was observed in the flexion- extension and the internal-external rotation of the shoulder joint and the flexion-extension of the elbow joint with respect to the grasp height, and the main effect which is significant difference was observed for all joints with respect to the inspection posture. For the interaction effect, there was a significant difference in the joints other than the flexion-extension of the wrist. Since the significant difference was observed in the interaction effect, the simple main effect test (Bonferroni’s test) was carried out. As a result, similar to the result of the muscle load, a characteristic tendency different from other inspection posture was observed in the inspection posture C. Specifically, the inspection posture C was different from other inspection postures at the abduction-adduction and inner-external rotation of the shoulder and the pronation-supination of the forearm, the joint displacement was the smallest at the eye level and the greatest at the downward from the eye level. In addition, the joint displacement in the inspection posture C increased as grasp height increased, however the tendency of other inspection postures was opposite. For the following parts, we discussed the main effect.

    3.2.1 Grasp Height

    In the flexion-extension of the shoulder and elbow joints, joint displacement increased at higher grasp heights. Since at high grasp heights the inspection object was far from the shoulder, subjects extended the elbow joint while bending the shoulder joint. On the other hand, in the internal-external rotation of the shoulder joint, joint displacement increased at lower grasp heights. Since at low grasp heights the inspection object was close to the shoulder, subjects internally rotated the shoulder joint while depressing the upper limb. In all other DOFs, the effect of joint displacement due to the difference in grasp height was small.

    3.2.2 Inspection Posture

    The various inspection postures influenced joint displacement for different DOFs in different ways as follows. In abduction-adduction of the shoulder joint, joint displacement increased in inspection posture B. Inspection posture B required subjects to make the forearm perpendicular at the front facing the body, thus adducting the shoulder joint. In flexion-extension of the shoulder and elbow joints, joint displacement increased in inspection posture C. Inspection posture C’s wrist position was higher than that of other inspection postures, thus extending the elbow joint while bending the shoulder joint. In pronation- supination of the forearm, joint displacement increased in inspection posture D. Joint displacement of the wrist joint was small in inspection posture D so that any orientation adjustments of the inspection object surface were accomplished by the supination of the forearm. In radial-ulnar flexion of the wrist joint, joint displacement increased in inspection posture C. Meanwhile, in palmardorsal flexion of the wrist joint, joint displacement increased in inspection posture F. In each case, this was because the inspection posture of note required any orientation adjustments of the inspection object surface to be accomplished by the movement of the wrist joint. In internal- external rotation of the shoulder joint, there was no significant joint displacement difference among inspection postures.

    3.2.3 Recommended Conditions

    Experimental results revealed that grasp height had a large effect on the proximal part of the upper limb, and the inspection posture had a large effect on the distal part of the upper limb. The cumulative joint displacement was found to be greater at higher grasp heights because, as the grasp height increases, the flexion angle of the shoulder joint also increases. Further, the cumulative joint displacement was found to be greatest with inspection posture C and smallest with inspection posture A. Considering each joint, inspection posture C caused the greatest displacements along one DOF of the shoulder joint, one DOF of the elbow joint, and one DOF of the wrist joint. This is because, in inspection posture C, it is necessary to extend the shoulder and elbow joints unlike with other inspection postures. Inspection posture A caused the lowest displacements along one DOF of the shoulder joint and one DOF of the forearm. This is because the adduction and abduction of the shoulder joint and the pronation and supination of the forearm are close to the neutral positions. From these analyses, we can conclude that inspection posture C is an undesirable working posture, and inspection posture A is the closest to a natural position.

    3.3 Subjective Evaluation

    Table 3 shows the means and standard deviations of the subjective evaluation metrics. In addition, Figure 21 shows the subjective physical load in the upper limb, and Figure 22 shows the perceived difficulty of assuming each posture. The asterisks in the graphs denote significant difference of 5%. Both subjective evaluations indicated higher scores (i.e., more difficulty) with inspection postures B and C and lower scores (i.e., lower difficulty) with inspection postures A and D. The correlation coefficient between the subjective physical load of the upper limb and the difficulty of assuming a posture was 0.94, indicating a strong correlation. The quantitative evaluations revealed that inspection posture D was associated with the least workload and inspection posture A was associated with the lowest cumulative joint displacement. Thus, the subjective evaluation results are consistent with the results regarding the muscle load and the cumulative joint displacement.

    3.4 Comprehensive Evaluation

    All evaluations indicate that the lowest workload on the upper limb occurs when the grasp height is below eye level. However, three evaluation indices offer different answers with regard to the best inspection posture to produce the least upper limb load: inspection posture D was the best in terms of muscle load, inspection posture A was the best in terms of working posture, and inspection postures A and D were both better than the other inspection postures based on the subjective evaluation. This difference can be attributed to the pronation-supination angle of the forearm (see Figure 23). With inspection posture D, the supination angle of the forearm is almost at the maximum possible value even though the muscle load is small. This large movement of the forearm should have had a negative effect on the working posture. In other words, inspection posture D is optimal if only the muscle load is evaluated, and inspection posture A is recommended after considering the loads on the joints due to the supination of the forearm, the subjective workload, and the difficulty of assuming the work posture.

    4. CONCLUSIONS

    In this study, we conducted an experiment to characterize the upper limb load during human visual inspection involving grasping an object using one hand. Combinations of three different grasp heights and six different inspection postures were evaluated by electromyography, joint angle measurements, and subjective evaluations. The workload was greater at higher grasp heights. However, different inspection postures were identified as the best depending on the evaluation results considered. Inspection posture D was found to be the best in terms of the muscle load, inspection posture A was found to be the best in terms of the working posture, and both were found to be more favorable than the other inspection postures in terms of the subjective evaluations. Therefore, in order to evaluate the workload associated with inspections and similar prolonged tasks with light loads, there will always be a need for subjective evaluations of the difficulty of assuming and maintaining various postures. By meaningfully integrating these subjective measures, we can effectively substantiate and successfully apply the results of data-driven ergonomic analysis.

    NOTICE

    This paper was reported at the Institute of Industrial Engineers Asian Conference 2013 (IIE 2013) as a proceeding and was expanded for that purpose.

    Figure

    IEMS-18-1-78_F1.gif

    Grasp height.

    IEMS-18-1-78_F2.gif

    Inspection posture.

    IEMS-18-1-78_F3.gif

    Visible range of symbol.

    IEMS-18-1-78_F4.gif

    Trapezius m (upper part).

    IEMS-18-1-78_F5.gif

    Pectoralis major m (clavicular part).

    IEMS-18-1-78_F6.gif

    Deltoid m (middle part).

    IEMS-18-1-78_F7.gif

    Biceps brachii m.

    IEMS-18-1-78_F8.gif

    Round pronator m.

    IEMS-18-1-78_F9.gif

    Flexor carpi ulnaris m.

    IEMS-18-1-78_F10.gif

    Extensor carpi radialis longus m.

    IEMS-18-1-78_F11.gif

    Thenar m.

    IEMS-18-1-78_F12.gif

    Integral workload of the upper limb muscles.

    IEMS-18-1-78_F13.gif

    Shoulder abduction-adduction.

    IEMS-18-1-78_F14.gif

    Shoulder flexion-extension.

    IEMS-18-1-78_F15.gif

    Shoulder internal-external.

    IEMS-18-1-78_F16.gif

    Elbow flexion-extension.

    IEMS-18-1-78_F17.gif

    Forearm pronation-supination.

    IEMS-18-1-78_F18.gif

    Wrist radial-ulnar flexion.

    IEMS-18-1-78_F19.gif

    Wrist palmar-dorsal flexion.

    IEMS-18-1-78_F20.gif

    Mean of cumulative joint displacement.

    IEMS-18-1-78_F21.gif

    Subjective physical load.

    IEMS-18-1-78_F22.gif

    Difficulty of assuming of the overall upper limb.

    IEMS-18-1-78_F23.gif

    Pronation-supination angle of forearm a posture.

    Table

    Mean and standard deviation of the %MVC for each muscle

    Mean and standard deviation of joint displacement for each DOF

    Mean and standard deviation of subjective evaluation

    REFERENCES

    1. Baudet, N. , Maire, J. L. , and Pillet, M.(2013), The visual inspection of product surfaces, Food Quality and Preference, 27(2), 153-160.
    2. Björkstén, M. and Jonsson, B. (1977), Endurance limit of force in long term intermittent static contractions, Scandinavian Journal of Work, Environment & Health, 3(1), 23-27.
    3. Cho, C. S. , Chung, B. M. , and Park, M. J.(2005), Development of real-time vision-based fabric inspection system, IEEE Transactions on Industrial Electronics, 52(4), 1073-1079.
    4. Criswell, E. (eds.)(2010), Cram’s Introduction to Surface Electromyography (2nd ed.), Jones and Bartlett Publishers, Massachusetts.
    5. Drury, C. G. (1973), The effect of speed of working on industrial inspection accuracy, Applied Ergonomics,4(1), 2-7.
    6. Garrett, S. K. , Melloy, B. J. , and Gramopadhye, A. K. (2000), The effects of per-lot and per-item pacing on inspection performance, International Journal of Industrial Ergonomics, 27(5), 291-302.
    7. Gramopadhye, A. K. , Bhagwat, S. , Kimbler, D., and Greenstein, J.(1998), The use of advanced technology for visual inspection training, Applied Ergonomics, 29(5), 361-375.
    8. Hida, T. and Seo, A. (2012), Development of evaluation function for estimating upper-limb load during visual inspection, Proceedings of the 13th Asia Pacific Industrial Engineering and Management Systems Conference, Phuket, Thailand, 1773-1778.
    9. Hida, T. , Chihara, T. , and Seo, A.(2013a), Effect of object size and grasp position on upper limb load during visual inspection, Journal of Japan Industrial Management Association, 63(4), 217-244 (in Japanese).
    10. Hida, T. , Chihara, T. , and Seo, A.(2013b), Effect of speed and scanning direction on upper limb load during visual inspection, Journal of Japan Industrial Management Association, 64(1), 1-8 (in Japanese).
    11. Hida, T. and Seo, A. (2015), Effect of inspection side and grasp position on upper limb load during visual inspection, Journal of Japan Industrial Management Association, 65(4), 260-267 (in Japanese).
    12. Jebaraj, D. , Tyrrell, R. A. , and Gramopadhye, A. K.(1999), Industrial inspection performance depends on both viewing distance and oculomotor characteristics, Applied Ergonomics, 30(3), 223-228.
    13. Kleiner, B. M. and Drury, C. G. (1993), Design and evaluation of an inspection training programme, Applied Ergonomics, 24(2), 75-82.
    14. Lee, F. C. and Chan, A. H. (2009), Effects of magnification methods and magnifier shapes on visual inspection, Applied Ergonomics, 40(3), 410-418.
    15. Liao, M. H. and Drury, C. G. (2000), Posture, discomfort and performance in a VDT task, Ergonomics, 43(3), 345-359.
    16. Lin, J. T. , Liang, G. F. , Hwang, S. L., and Wang, E. M.(2010), Predicting visual fatigue in integrated circuit packaging plants, Human Factors and Ergonomicsin Manufacturing & Service Industries, 20(5), 461-469.
    17. McAtamney, L. and Corlett, E. N. (1993), RULA: A survey method for the investigation of work-related upper limb disorders, Applied Ergonomics, 24(2), 91-99.
    18. Megaw, E. D. and Richardson, J. (1979), Eye movements and industrial inspection, Applied Ergonomics, 10(3), 145-154.
    19. Melloy, B. J. , Harris, J. M. , and Gramopadhye, A. K.(2000), Predicting the accuracy of visual search performance in the structural inspection of aircraft, International Journal of Industrial Ergonomics, 26(2), 277-283.
    20. Nakajima, R. , Asano, Y., Hida, T. , and Matsumoto, T.(2018), The relationship between dirt levels of inspection surface and defect detection in visual inspection utilizing peripheral vision, Industrial Engineering & Management Systems, 17(1), 102-112.
    21. Nalanagula, D. , Greenstein, J. S. , and Gramopadhye, A. K.(2006), Evaluation of the effect of feedforward training displays of search strategy on visual search performance, International Journal of Industrial Ergonomics, 36(4), 289-300.
    22. Perotto, A. O. (eds.)(1994), Anatomical Guide for the Electromyographer: The Limbs and Trunk (3rd ed.), Charles C Thomas Pub Ltd, Illinois.
    23. Schneider, J. , Kluth, K. , and Strasser, H.(2013), Development and evaluation of specific lighting scenarios and ergonomic approaches to optimise visual inspection tasks, Theoretical Issues in Ergonomics Science, 14(2), 195-211.
    24. See, J. E. (2015), Visual inspection reliability for precision manufactured parts, Human Factors, 57(8), 1427-1442.
    25. Yeow, P. H. and Sen, R. N. (2004), Ergonomics improvements of the visual inspection process in a printed circuit assembly factory, International Journal of Occupational Safety and Ergonomics, 10(4), 369-385.