Search for a command to run...
Driver inattention and distraction are the main causes of road accidents, many of which result in fatalities. To reduce road accidents, the development of information systems to detect driver inattention and distraction is essential. Currently, distraction detection systems for road vehicles are not yet widely available or are limited to specific causes of driver inattention such as driver fatigue. Despite the increasing automation of driving due to the availability of increasingly sophisticated assistance systems, the human driver will continue to play a longer role as supervisor of vehicle automation. With this in mind, we review the published scientific literature on driver distraction detection methods and integrate the identified approaches into a holistic framework that is the main contribution of the paper. Based on published scientific work, our driver distraction detection framework contains a structured summary of reviewed approaches for detecting the three main distraction detection approaches: manual distraction, visual distraction, and cognitive distraction. Our framework visualizes the whole detection information chain from used sensors, measured data, computed data, computed events, inferred behavior, and inferred distraction type. Besides providing a sound summary for researchers interested in distracted driving, we discuss several practical implications for the development of driver distraction detection systems that can also combine different approaches for higher detection quality. We think our research can be useful despite - or even because of - the great developments in automated driving.