PHYSICAL TASKS:WORKPLACE ANALYSIS AND DESIGN

WORKPLACE ANALYSIS AND DESIGN

Evaluation of Working Postures

The posture of the human body at work is influenced by several factors, including workstation layout (heights of the workplace, orientation of tools and work objects), hand tool design, work methods and work habits, visual control and force exertion requirements, and anthropometric characteristics of the workers (Chaffin et al. 1999; Grandjean 1980; Habes and Putz-Anderson, 1985; Corlett et al. 1986; Kilbom et al. 1986; Keyserling 1986; Wallace and Buckle 1987). Poor and unnatural (i.e., not- neutral) working postures have been associated with the onset of fatigue, bodily discomforts and pains, and musculoskeletal disorders (Tichauer 1978; Karhu et al. 1977; and Keyserling et al. 1988). Keyserling (1990) discusses the scientific evidence of such associations. For example, it was shown that trunk flexion, lateral bending, or twisting increases muscle stress and intervetrabral disc pressure, while prolonged sitting or forward bending leads to increased risk of low back pain and muscle fatigue (Chaffin 1973; Schultz et al. 1982; Kelsey and Hochberg 1988). Prolonged elevation of the arms may cause tendonitis (Hagberg 1984), while shoulder extension and elevation may lead to thoracic outlet syndrome (Armstrong 1986). Also, strong association was found between poor neck posture and cervicobrachial disorders (Jonsson et al. 1988).

Computer-Aided Analysis of Working Postures

Ergonomics provides useful guidelines for evaluation of working postures, especially with respect to identification and quantification of postural stresses and their relation to posture-related work injury. The ultimate goal of such analysis is to improve the workplace design by reducing postural stresses imposed upon the body to the acceptable (safe) levels. Some of the methods used in the past to systematically evaluate work postures by using computerized or semicomputerized techniques are reported by Karhu et al. (1977); Corlett et al. (1979); Holzmann (1982); Keyserling (1986); Pearcy et al. (1987); and Wangenheim and Samuelson (1987). Snijders et al. (1987) introduced devices for measurement of forward bending, lateral bending, and torsion continuously. Ferguson et al. (1992) used a lumbar motion monitor to measure the back motion during asymmetric lifting tasks. The Ovako Working Posture Analysis System (OWAS), which uses predefined standard postures, was first reported by Karhu et al. (1977). The posture targeting technique (1988) and RULA (1996), developed by Corlett et al. (1979), are based on the recording of the positions of the head, trunk, and upper and lower arms.

Postural Analysis Systems

OWAS

OWAS (the Ovako Working Posture Analyzing System), first reported by Karhu et al. (1977), iden- tifies the most common work postures for the back, arms, and legs and estimates the weight of the loads handled or the extent of the strength (effort). A rating system categorizes 72 different postures in terms of discomfort caused and the effect on health. Back postures are defined as either straight, bent, straight and twisted, or bent and twisted. No specificity (in terms of number of degrees) is provided. This categorization results in the specification of four action categories. The observed posture combinations are classified according to the OWAS method into ordinal scale action cate- gories. The four action categories described here are based on experts’ estimates on the health hazards of each work posture or posture combination in the OWAS method on the musculoskeletal system:

1. Work postures are considered usually with no particular harmful effect on the musculoskeletal system. No actions are needed to change work postures.

2. Work postures have some harmful effect on the musculoskeletal system. Light stress, no im- mediate action is necessary, but changes should be considered in future planning.

3. Work postures have a distinctly harmful effect on the musculoskeletal system. The working methods involved should be changed as soon as possible.

4. Work postures have an extremely harmful effect on the musculoskeletal system. Immediate solutions should be found to reduce these postures.

OWASCA, a computer-aided visualizing and training software for work posture analysis, was developed using OWAS. OWASCA is intended as OWAS training software (Vayrynen et al. 1990).

The system is also suitable for visualizing the work postures and for the basic analysis of the postures and their loads. The posture is presented with parametric vector using 2D graphics, OWAS codes, and texts. The posture of the back, arms and legs, posture combination, force or effort used, additional postures, and action categories can be studied interactively step by step. The required OWAS skills can be tested by OWASCA. The program shows a random work posture, and the user is asked to identify it. OWASCA describes the errors and gives the numbers of test postures and correct answers (Mattila et al. 1993).

Standard Posture Model

A standard system for analyzing and describing postures of the trunk, neck, shoulders, and lower extremities during dynamic work was developed at the University of Michigan (Keyserling 1990) (see Figure 10). Neutral joint postures and their deviations were also defined (Table 12). The postural

analysis involves three steps. First, a continuous video recording of the job is obtained. Second, a sequential description of the major tasks required to perform the job is done in a laboratory, with the job being broken into fundamental work elements and their times measured. The third and final step involves collection of the postural data using the common time scale developed from the fun- damental work elements. Postural changes are keyed into the system through the preassigned keys corresponding to specific postures. The value of each posture and the time of postural change for a given joint are recorded and stored in a computer. Based on the above data, the system generates a posture profile for each joint, consisting of the total time spent on each standard posture during the work cycle, the range of times spent in each standard posture, the frequency of posture use, and so on. The system can also provide a graph showing postural changes over time for any of the body joints (segments) of interest.

Acceptability of Working Postures

Analysis of posture must take into consideration not only the spatial elements of the posture, that is, how much is the person flexed, laterally bent, or rotated (twisted), but how long these postures are maintained. Milner et al. (1986) pointed out where an individual is working to the limits of endurance capacity, it has been found that full recovery is not possible within a rest period 12 times the maximum holding time. Full recovery is possible as long as the holding time is a small percentage of maximum handling time.

Bonney et al. (1990) studied tolerability of certain postures and showed that complex postures requiring movement in more than one direction are more uncomfortable than simple postures. Lateral bending produced more discomfort than either flexed or rotated postures and appears to be the least well-tolerated posture. Rotation by itself does not cause significant discomfort. This finding is con- sistent with epidemiological results of Kelsey and Golden (1988), who hypothesized that twisting along may not produce enough stress to bring about a detectable increase in risk.

Corlett and Manenica (1980) derived estimates for maximum handling times for various postures when performing a no-load task. These recommendations are as follows:

1. Slightly bent forward postures (approx. 15–20°) 8 min

2. Moderately bent forward posture (approx. 20–60 °) 3–4 min

3. Severely bent forward postures (greater than about 60°) approx. 2 min

Colombini et al. (1985) presented criteria on which posture assessments should be based. Postures considered tolerable include (1) those that do not involve feelings of short-term discomfort and (2) those that do not cause long-term morpho-functional complaints. Short-term discomfort is basically the presence of a feeling of fatigue and / or pain affecting any section of the asteo-arthromuscular and ligamentous apparatus appearing in periods lasting minutes, hours, or days.

Miedema et al. (1997) derived the maximum holding times (MHT) of 19 standing postures in terms of percent of shoulder height and percent of arm reach. They also classified such working postures into three categories, depending on the mean value of the MHT: (1) comfortable; (2) mod- erate; and (3) uncomfortable postures (see Table 13).

Recently, Kee and Karwowski (2001) presented data for the joint angles of isocomfort (JAI) for the whole body in sitting and standing postures, based on perceived joint comfort measures. The JAI value was defined as a boundary indicating joint deviation from neutral (0°), within which the per- ceived comfort for different body joints is expected to be the same. The JAI values were derived for nine verbal categories of joint comfort using the regression equations representing the relationships between different levels of joint deviation and corresponding comfort scores for each joint motion. The joint angles with marginal comfort levels for most motions around the wrist, elbow, neck, and ankle were similar to the maximum range-of-motion (ROM) values for these joints. However, the isocomfort joint angles with the marginal comfort category for the back and hip motions were much smaller than the maximum ROM values for these joints.

There were no significant differences in percentage of JAI in terms of the corresponding maximum ROM values between standing and sitting postures. The relative marginal comfort index, defined as the ratio between joint angles for marginal comfort and the corresponding maximum ROM values, for hip was the smallest among all joints. This was followed in increasing order of the index for lower back and for shoulder, while the index values for elbow were the largest. This means that hip motions are less comfortable than any other joint motion, while elbow motions are the most com- fortable. The relative good comfort index exhibited much smaller values of joint deviation, with most

index values of less than 40.0. The presented data about joint angles of isocomfort can be used as design guidelines for postural comfort in a variety of human–machine systems.

International Standards

A list of international standards in the area of anthropometry and biomechanics being developed by ISO is shown in Table 14.

Exposure Assessment of Upper-Limb Repetitive Movements

Recently, Colombini et al. (1999) reported the findings of an international expert group working under auspices of the Technical Committee on Musculoskeletal Disorders of the International Ergon- omics Association (IEA) and endorsed by the International Commission on Occupational Health (ICOH). This report provides a set of definitions, criteria, and procedures for assessment of working conditions with respect to exposure of the upper extremities. The document includes two important international standards: Evaluation of Working Postures (ISO / DIS 11226 1998), presented in Figure 11, and Evaluation of Working Postures in Relation to Machinery (CEN prEN 1005—4, 1997): Upper arm elevation, shown in Figure 12.

European Standards for Working Postures During Machinery Operation The draft proposal of the European (CEN / TC122, WG4 / SG5) and the international (ISO TC159 / SC3 / WG2) standardization document (1993) ‘‘Safety of machinery—human physical performance, Part 4: Working postures during machinery operation,’’ specifies criteria of acceptability of working postures vs. exposure times. These criteria are discussed below.

General Design Principles Work task and operation should be designed with sufficient physical and mental variation so that the physical load is distributed over various postures and patterns of movements. Designs should accommodate the full range of possible users. To evaluate whether working postures during machinery operation are acceptable, designers should perform a risk as- sessment, that is, an evaluation of the actual low-varying (static) postures of body segments. The lowest maximum acceptable holding time for the various body segment postures should be deter- mined.

Assessment of Trunk Posture An asymmetric trunk posture should be avoided (no trunk axial rotation or trunk lateral flexion). Absence of normal lumbar spine lordosis should be avoided. If the trunk is inclined backward, full support of the lower and upper back should be provided. The forward trunk inclination should be less than 60°, on the condition that the holding time be less than the maximum acceptable holding time for the actual forward trunk inclination, as well as that adequate rest is provided after action (muscle fitness should not be below 80%).

Assessment of Head Posture An asymmetric head posture should be avoided (no axial rotation or lateral flexion of the head with respect to the trunk). The head inclination should not be less than trunk inclination (no neck extension). The head inclination should not be larger than the trunk inclination for more than 25° (no extreme neck flexion). If the head is inclined backward, full head support should be provided. The forward head inclination should be less than 25° (if full trunk support is provided), forward inclination should be between less than 85°, on the condition that the holding time should be less than the maximum acceptable holding time for the actual forward head inclination as well as that adequate rest is provided.

Assessment of Upper Extremity Posture Shoulder and upper arm posture. Upper arm retroflexion and upper-arm adduction should be avoided. Raising the shoulder should be avoided. The upper-arm elevation should be less than 60°, on the condition that the holding time be less than the maximum acceptable holding time for the actual upper-arm elevation as well as that adequate rest be provided after action (muscle fitness should not be below 80%).

Forearm and hand posture. Extreme elbow flexion or extension, extreme forearm pronation or supination, and extreme wrist flexion or extension should be avoided. The hand should be in line with the forearm (no ulnar / radial deviation of the wrist).