Knowledge elicitation techniques for modelling intentional systems with Cognitive Work Analysis

The design of new human machine systems, and the analysis of current human machine systems pose interesting dilemmas for the human factors specialist. Cognitive Work Analysis (CWA), (Vicente, 1999; Rasmussen, Pejtersen, & Goodstein, 1994) is an analytical framework for modeling both work environments and safe, flexible strategies workers can use to act upon this environment. CWA has proven to be a particularly effective modeling tool not only for design, but also throughout the entire system design life cycle (Sanderson, Naikar, Lintern & Goss, 2000). An important consideration for CWA is the human requirements of a new system, or the evaluation of how humans interact with existing systems. In this paper we describe four knowledge elicitation (KE) techniques developed over the last year to assist in the rapid development of human-centered CWA based analyses. These KE techniques were developed for work in two different aviation domains, an existing civilian Search and Rescue (SAR) service, and a new airborne defence platform called Airborne Early Warning and Control (AEW&C) which is yet to exist physically. A common link between the SAR domain and the AEW&C domain is that they are both systems run largely through rules of practice that contain a ‘domain of potential risk’ for the workers involved. In the case of SAR the domain of risk can be thought of as ‘passive’, in that there are no malicious intentions (i.e. simply a patient lying in the water waiting to be rescued). However in the case of AEW&C, the domain of risk can be thought of as “active” in that there is a intelligent adversary with their own intentions and beliefs about the state of the environment. The appropriateness of the KE techniques will be discussed; a) with respect to the two aviation domains and b) with respect to the use of CWA products across different stages of the system design life cycle.