Automated-Manual Transitions: Human Capabilities and Adaptive Cruise Control

ITS innovations in California are likely to include automated systems for vehicle guidance. Such systems will supplant manual controls during certain types of vehicle operation. However, the alternative manual control must remain intact in the vehicle. Thus, epochs of automated-manual transition (A-MT) are inevitable. The problem is how to characterize a given transition type and then how to optimize it. In this study we examined one of the several predictable transitions that use of the Adaptive Cruise Control (ACC) will lead to. This predictable A-MT is the event that would ensue when a lead vehicle (LV) suddenly brakes maximally and a following vehicle (FV) under the control of ACC must react appropriately (hereafter referred to as the LV Braking Scenario). Such an event is neither unlikely nor is it benign. It can be shown that within the current design of ACC, this event will lead to a collision. Collision avoidance is possible but only if the human operator (HO) of the FV assumes manual control in a timely way. In TO 4221 the conditions required for a graceful A-MT in this scenario were investigated. We focused attention on two features of the HO, those visual capabilities required by the HO to determine the need to assume manual control and also those features of an in-vehicle warning signal (initiated by either vehicle) that could reliably prompt appropriate HO action. This was accomplished in the context of three tasks.