Effectiveness of Workload-Based Drowsy Driving Countermeasures 2019-01-1228
This study evaluated the effectiveness of alternative workload-based interventions intended to restore driver alertness following drowsy episodes. Unlike traditional drowsy driving studies, this experiment did not target sleep-deprived individuals, but rather studied normally rested drivers under the assumption that low-workload environments could trigger drowsy driving episodes. The study served as a proof of concept for varying the nature and onset of countermeasure interventions intended to disrupt the drowsiness cycle. Interventions to combat drowsiness attempted to target driver workload, either physical or cognitive, and included two primary treatment conditions: 1) physical workload to increase driver steering demands and 2) trivia-based interactive games to mentally challenge drivers. A benchmark comparison condition using music was also investigated to contrast the relative influence of workload-based interventions with passive listening to musical arrangements. The study also varied the onset stage of the intervention, basing either early or late onset on driver drowsiness levels indexed using a Percentage of Eyelid Closure (PERCLOS) measure. Thirty drivers, aged 21–70, completed a 3-hour trip in a driving simulator. When a drowsy driving episode was identified, the driver received the prescribed countermeasure for a fixed time period. The study method successfully induced multiple drowsy driving episodes of varying magnitudes and durations during the simulated trips. Results suggest that both physical and cognitive workload-based countermeasures can effectively combat drowsiness and reengage drivers. Increasing the physical workload via steering demands and the cognitive workload via gaming interactions were equally effective at restoring drivers to an alert state following a drowsy episode, with average effectiveness levels of 98% and 87%, respectively. In contrast, listening to music was less effective, restoring drivers to an alert state about 68% of the time.
Freddy V. Rayes, Maureen Short, Jason Meyer, Robert E. Llaneras
General Motors, Virginia Tech Transportation Institute, Virginia Polytechnic Inst. & State Univ.