Automatic detection of driver's awareness with cognitive task from driving behavior

Awareness or alertness while driving is extremely important for road safety. It is known that cognitive load due to multitasking affects driving alertness. In this work, possibilities of automatic detection of driver's distraction due to various cognitive tasks from driving behavior has been studied with simulation experiments in a driving simulator. Experiments are done with different types of driving situations (normal driving, driving with secondary cognitive tasks) in different road scenarios. Various time series data (such as steering wheel angle, brake stroke, accelerator stroke, car speed etc.) from sensors attached to car simulator have been analyzed with data mining algorithms. It has been found that driving behavior changes with statistical significance for varying cognitive tasks and also the change of driving behavior of individual person is reflected in different parameters of driving behavior.