Although simulators have been widely used as training environments in different industries (e.g. oil and nuclear power), there is little rigorous empirical work evaluating the effectiveness of the training methods employed. This article examines the use of simulator training in process control environments. The results of an exploratory field study are reported and the current practices of simulator training are described. The study revealed that simulator training varied considerably across organisations, often with little theoretical or empirical work to guide training design. To evaluate the utility and effectiveness of different methods of simulator training in process control environments, the article also presents a literature review of the research on laboratory-and field-based training. Several training methods are identified as having particular potential for temporal and adaptive transfer and are to be empirically tested in future studies.

In complex work environments, the occurrence of novel system states represents a particular challenge for the design of training. This article is concerned with the use of heuristic rules to prepare operators for the management of unfamiliar fault states. An experiment was carried out to examine the effects of heuristic rule training on operator performance and system management behaviour. Thirty-nine trainee operators from the chemical industry took part in the study. They were trained for 4 h on a PC-based simulation of a process control task. Operators in the experimental group received training on heuristic rules while operators in the control group did not. One week later the operators participated in a 70-min testing session. While the results showed that heuristic rules training led to better diagnostic performance, it was also associated with increased operator fatigue and impaired secondary task performance. The implications of the findings for using heuristic rule training are discussed.