human factors

Aviation Maintenance 26-9
8. Human-factors guide for aviation maintenance. With so much research and development activities
on human factors in maintenance and inspection, there is an obvious need to get usable information
for the nonspecialists within the system. Since 1992, a guide has been under development to
codify the human-factors principles, techniques, and findings for the system participants, such as
managers and supervisors of maintenance and inspection. This guide was produced in CDROM
and hard copy forms in the mid-1990s (Maddox, 1995), and has formed the basis of training and
interventions in the industry.
9. Workcard redesign. As existing workcards were often found to be unsatisfactory from a humanfactors
viewpoint, a project was undertaken to show how they could be improved. The first phase
of this project (Patel, Drury, & Prabhu, 1993) used the human-factors literature to the determine
principles of information design applicable to workcards, and to design new workcards embodying
these principles. These new workcards were developed as job aids for two distinct types of
inspection. For a C-check, which is a heavy inspection conducted infrequently, inspectors need
detailed guidance on what defects to expect and which areas to search. For the more frequent
A-checks, the inspection is typically the same every day (or more accurately, every night), and
hence, a layered information system is needed. Here, a checklist provides procedural and sequence
information to prevent procedural errors, and more detailed information is available behind the
checklist for reference as needed. Evaluation of the C-check prototype showed highly significant
improvements when the inspectors rated the workcard design (Patel, Drury, & Lofgren, 1994).
Since that study, there has been much interest in using AECMA Simplified English for workcards.
Chervak, Drury, and Ouellette (1996) showed that Simplified English did reduce comprehension
errors. Later, Chervak and Drury (2003) demonstrated that maintenance errors were also
reduced in a simple maintenance task. A design tool for workcards (documentation design aid
[DDA]) was developed by Drury and Sarac (1997), and has been used extensively (available for
download at http://hfskyway.faa.gov). A more recent study of the DDA in outsourced aviation
maintenance (Drury, Wenner, & Kritkausky, 1999) showed that it could reduce half of the comprehension
errors. Currently, Simplified English is one of the several techniques being tested in overseas
repair stations, as a defense against language errors in maintenance (Drury & Ma, 2004).
10. Restrictive space changes. Many aircraft inspection tasks must be performed in restrictive spaces
owing to airframe structural constraints. A study at an airline partner measured the effect of
restrictions on postural accommodations (e.g., movements), perceived discomfort, and perceived
workload (TLX). It was found that it is possible to differentiate between good and poor workspaces
using these measures, and to use the findings to initiate countermeasures in the form of
improved access equipment (Reynolds, Drury, & Eberhardt, 1994). A classification scheme for
restricted spaces was developed to assist this work, and was tested using laboratory simulations of
inspection tasks (Reynolds, Drury, Sharit, & Cerny, 1994).
11. Visual-search training. A comprehensive series of projects used a workstation-based visual
inspection simulator (Latorella et al., 1992) to test the various hypotheses about improvement
of inspection training. For visual-search training, both improvements in defect conspicuity and
improvements in search strategy were sought (Drury & Gramopadhye, 1992). Current inspection-training
procedures are largely either classroom-based, covering theory and regulation, or
on-the-job practice. Neither technique is most appropriate to the skills required in inspection,
particularly, the search skills. One experiment tested a technique of practice on a visual-lobe
testing task and showed that this practice transferred to search the performance for both similar
and perceptually similar defects. The second experiment evaluated both performance feedback
and cognitive feedback as techniques for improving search strategy and performance. It was
found (Drury & Gramopadhye, 1992) that the two types of feedback have different effects, and
hence, both may be needed to obtain the best results.