A Predictor-Compensator Design to Assist Human Decision-Making Process in an Air-Traffic-Control Simulator

Air traffic control is a demanding task for human operators, as this task requires tracking multiple events, managing the events, and taking actions in the presence of multiple and possibly competing objectives. In such critical tasks, human intelligence is extremely crucial however human decisions also become more prone to errors, which could cause tragic events. One idea to prevent such errors is to design smart machines that can assist human subjects in making decisions whenever human errors become more likely. In this article, we present a simulation model that captures the essence of how a human subject model would interact with a simplified version of an air traffic control simulator, and show how we design a predictor-compensator in order to regulate and possibly improve this interaction, such that overall human-machine interface can be optimized, and human workload is reduced on average.