LEVELS OF AUTOMATION: IMPACTS ON PERFORMANCE, WORKLOAD, AND SITUATION AWARENESS

摘要

Advanced reactors are expected to use higher degrees of automation than the existing fleet of light water reactors in the United States. Reliable automation tends to increase system performance, but often has negative human performance consequences (such as complacency, human-out-of-the-loop problems, etc.). Many researchers advocate using intermediate levels of automation to handle the system performance and human performance tradeoffs. However, this strategy represents a compromise that doesn't fully take advantage of the human's or the automation's capabilities. Researchers have encouraged the use of adaptive automation (defined as flexible or dynamic allocation of functions to humans and automatic agents) as way to employ higher levels of automation without sacrificing human performance (particularly situation awareness). Recently, researchers have demonstrated a type of adaptive automation that allows the operator to delegate tasks to automation in much the same way as he would to a human team. Researchers have conducted an experimental study comparing performance and situation awareness for four levels of automation using a simplified process control simulation. The study compared performance using manual, intermediate, adaptable, and fully automatic levels of automation. The researchers hypothesized that the adaptable automation would yield the best performance and SA combination, but found there was no difference between intermediate and adaptable automation. The results are discussed in the context of previous literature on human-automation interaction and in application to advanced reactors.