HUMAN FACTORS AUDIT:AUDIT SYSTEMS IN PRACTICE

AUDIT SYSTEMS IN PRACTICE

Almost any of the audit programs and checklists referenced in previous sections give examples of their use in practice. Only two examples will be given here, as others are readily accessible. These examples were chosen because they represent quite different approaches to auditing.

Auditing a Decentralized Business

From 1992 to 1996, a major U.S.-based apparel manufacturer had run an ergonomics program aimed primarily at the reduction of workforce injuries in backs and upper extremities. As detailed in Drury et al. (1999), the company during that time was made up of nine divisions and employed about 45,000 workers. Of particular interest was the fact that the divisions enjoyed great autonomy, with only a small corporate headquarters with a single executive responsible for all risk-management activities. The company had grown through mergers and acquisitions, meaning that different divisions had different degrees of vertical integration. Hence, core functions such as sewing, pressing, and distribution were common to most divisions, while some also included weaving, dyeing, and em- broidery. In addition, the products and fabrics presented quite different ergonomic challenges, from delicate undergarments to heavy jeans to knitted garments and even luggage.

The ergonomics program was similarly diverse. It started with a corporate launch by the highest- level executives and was rolled out to the divisions and then to individual plants. The pace of change was widely variable. All divisions were given a standard set of workplace analysis and modification tools (based on Drury and Wick 1984) but were encouraged to develop their own solutions to prob- lems in a way appropriate to their specific needs.

Evaluation took place continuously, with regular meetings between representatives of plants and divisions to present results of before-and-after workplace studies. However, there was a need for a broader audit of the whole corporation aimed at understanding how much had been achieved for the multimillion-dollar investment, where the program was strong or weak, and what program needs were emerging for the future. A team of auditors visited all nine divisions, and a total of 12 plants spread across eight divisions, during 1995. This was three years after the initial corporate launch and about two years after the start of shop-floor implementation.

A three-part audit methodology was used. First, a workplace survey was developed based on elements of the program itself, supplemented by direct comparisons to ergonomics standards and good practices. Table 9 shows this 50-item survey form, with data added for the percentage of ‘‘yes’’ answers where the responses were not measures or scale values. The workplace survey was given at a total of 157 workplaces across the 12 plants. Second, a user survey (Table 10) was used in an interview format with 66 consumers of ergonomics, typically plant managers, production managers, human resource managers, or their equivalent at the division level, usually vice presidents. Finally, a total of 27 providers of ergonomics services were given a similar provider survey (Table 11) interview. Providers were mainly engineers, with three human resources specialists and one line supervisor. From these three audit methods the corporation wished to provide a time snapshot of how effectively the current ergonomics programs was meeting their needs for reduction of injury costs. While the workplace survey measured how well ergonomics was being implemented at the workplace, the user and provider surveys provided data on the roles of the decision makers beyond the workplace.

Detailed audit results are provided in Drury et al. (1999), so only examples and overall conclusions are covered in this chapter. Workplaces showed some evidence of good ergonomic practice, with generally satisfactory thermal, visual, and auditory environments. There were some significant dif- ferences ( p < 0.05) between workplace types rather than between divisions or plants; for example, better lighting (> 700 lux) was associated with inspection and sewing. Also, higher thermal load was associated with laundries and machine load / unload. Overall, 83% of workplaces met the ASH- RAE (1990) summer comfort zone criteria. As seen in Table 12, the main ergonomics problem areas were in poor posture and manual materials handling. Where operators were seated (only 33% of all workplaces) seating was relatively good. In fact, many in the workforce had been supplied with well- designed chairs as part of the ergonomics program.

To obtain a broad perspective, the three general factors at the end of Table 9 were analyzed. Apart from cycle time (W48), the questions related to workers having seen the corporate ergonomics video (W49) and having experienced a workplace or methods change (W50). Both should have received a ‘‘yes’’ response if the ergonomics program were reaching the whole workforce. In fact, both showed highly significant differences between plants (X 2 = 92.0, p < 0.001, and X 2 = 22.2, p < 0.02, respectively). Some of these differences were due to two divisions lagging in ergonomics implemen- tation, but even beyond this were large between-plant differences. Overall, 62% of the workforce had seen the ergonomics video, a reasonable value but one with wide variance between plants and divi- sions. Also, 38% of workplaces had experienced some change, usually ergonomics-related, a respect- able figure after only two to three years of the program.

From the user and provider surveys an enhanced picture emerged. Again, there was variability between divisions and plants, but 94% of the users defined ergonomics as fitting the job to the operator rather than training or medical management of injuries. Most users had requested an ergonomic intervention within the past two months, but other ‘‘users’’ had never in fact used ergonomics.

The solutions employed ranged widely, with a predominance of job aids such as chairs or standing pads. Other frequent categories were policy changes (e.g., rest breaks, rotation, box weight reduction) and workplace adjustment to the individual operator. There were few uses of personal aids (e.g. splints) or referrals to MDs as ergonomic solutions. Changes to the workplace clearly predominated over changes to the individual, although a strong medical management program was in place when required. When questioned about ergonomics results, all mentioned safety (or workplace comfort or ease of use), but some also mentioned others. Cost or productivity benefits were the next most common response, with a few additional ones relating to employee relations, absence / turnover, or job satisfaction. Significantly, only one respondent mentioned quality.

The major user concern at the plant level was time devoted to ergonomics by providers. At the corporate level, the need was seen for more rapid job-analysis methods and corporate policies, such as on back belts or ‘‘good’’ chairs. Overall, 94% of users made positive comments about the ergon- omics program.

Ergonomics providers were almost always trained in the corporate or division training seminars, usually near the start of the program. Providers’ chief concern was for the amount of time and resources they could spend on ergonomics activities. Typically, ergonomics was only one job re- sponsibility among many. Hence, broad programs, such as new chairs or back belts, were supported enthusiastically because they gave the maximum perceived impact for the time devoted. Other so- lutions presented included job aids, workplace redesign (e.g., moving from seated to standing jobs for long-seam sewing), automation, rest breaks, job rotation, packaging changes, and medical man-

Any ergonomics changes to workplace or methods

agement. Specific needs were seen in the area of corporate or supplier help in obtaining standard equipment solutions and of more division-specific training. As with users, the practitioners enjoyed their ergonomics activity and thought it worthwhile.

Recommendations arising from this audit were that the program was reasonably effective at that time but had some long-term needs. The corporation saw itself as an industry leader and wanted to move beyond a relatively superficial level of ergonomics application. To do this would require more time resources for job analysis and change implementation. Corporate help could also be provided in developing more rapid analysis methods, standardized video-based training programs, and more standardized solutions to recurring ergonomics problems. Many of these changes have since been implemented.

On another level, the audit was a useful reminder to the company of the fact that it had incurred most of the up-front costs of a corporate ergonomics program, and was now beginning to reap the benefits. Indeed, by 1996, corporate injury costs and rates had decreased by about 20% per year after

peaking in 1993. Clearly, the ergonomics program was not the only intervention during this period, but it was seen by management as the major contributor to improvement. Even on the narrow basis of cost savings, the ergonomics program was a success for the corporation.

Error Reduction at a Colliery

In a two-year project, reported by Simpson (1994) and Fox (1992), the human error audit described in Section 3.2 was applied to two colliery haulage systems. The results of the first study will be presented here. In both systems, data collection focused on potential errors and the performance- shaping factors (PSFs) that can influence these errors. Data was collected by ‘‘observation, discussion and measurement within the framework of the broader man-machine systems and checklist of PSFs,’’ taking some 30–40 shifts at each site. The whole haulage system from surface operations to delivery at the coal face was covered.

The first study found 40 active failures (i.e., direct error precursors) and nine latent failures (i.e., dormant states predisposing the system to later errors). Four broad classes of active failures were:

1. Errors associated with locomaintenance (7 errors), e.g., fitting incorrect thermal cut-offs

2. Errors associated with locooperation (10 errors), e.g., locos not returned to service bay for 24- hour check.

3. Errors associated with loads and load security (7 errors); e.g., failure to use spacer wagons between overhanging loads

4. Errors associated with the design / operation of the haulage route (10 errors), e.g., continued use despite potentially unsafe track

5. Plus a small miscellaneous category The latent failures were (Fox 1992):

1. Quality assurance in supplying companies

2. Supplies ordering procedures within the colliery

3. Locomotive design

4. Surface make-up of supplies

5. Lack of equipment at specific points

6. Training

7. Attitudes to safety

8. The safety inspection / reporting / action procedures

As an example from 3, Locomotive design, the control positions were not consistent across the locomotives fleet, despite all originating from the same manufacturer.

Using the slip / mistake / violation categorization, each potential error could be classified so that the preferred source of action (intervention) could be specified.

This audit led to the formation of two teams, one to tackle locomotive design issues and the other for safety reporting and action. As a result of team activities, many ergonomic actions were imple- mented. These included management actions to ensure a uniform wagon fleet, autonomous inspection/ repair teams for tracks, and multifunctional teams for safety initiatives.

The outcome was that the accident rate dropped from 35.40 per 100,000 person-shifts to 8.03 in one year. This brought the colliery from worst in the regional group of 15 collieries to best in the group, and indeed in the United Kingdom. In addition, personnel indicators, such as industrial rela- tions climate and absence rates, improved.