Mental Representations as Mirror of Function Allocation and Control Strategies in a Simulated Chemical Plant

Abstract

In order to fulfill supervisory control tasks in automated systems, a human operator needs to have an adequate mental representation of the process and the technical system. The development of such representations depends on both the technically provided control opportunities, as determined by the allocation of functions between the technical system and the human operator, and the use the human operator makes of such opportunities as expressed in the chosen control strategies in process monitoring, failure diagnosis and process intervention. At the same time, operators' mental representations of a system influence the choice of control strategies.

A study will be presented that aims at describing this relationship between function allocation, control strategies and mental representations in a quasi-experimental setting. Apprentices of a Swiss chemical company operated a simulated plant with varying degrees of automation and with process disturbances of varying severity. Observations and interviews were used to obtain data on control strategies, while the apprentices' mental representations of the process and the technical system were inferred from written process protocols and structural diagrams drawn after the experiment.