Performance measurement paradigms for driving simulator research

Performance measures can be crucial to research objectives. Performance measurement is central to tracking and documenting training progress, and providing objective quantification of experimental treatments. Performance measurement can also afford criteria for diagnostic assessment procedures concerning driver capability, qualification and certification. The roadway environment presents a variety of situations that require drivers to control steering and speed, and decide between alternative actions. Traffic signals provide one such common occurrence where drivers must decide whether or not to stop. Another typical situation involves encounters with unexpected obstacles. Following roadway profiles and negotiating construction zones results in more continuous control of steering and speed. This paper presents models for decision/reaction/maneuver timing, and for continuous steering and speed control. The models are applied to driver behavior obtained through real-world observation, and experiments employing laboratory simulation and instrumented vehicle tests. The requirements for structuring driving scenarios to accomplish the measurement paradigms are discussed.