Towards multi-modal wearable driver monitoring: Impact of road condition on driver distraction

The objective of this paper is to propose initial steps towards the design of the next generation multi-modal driver monitoring platform to be facilitated in urban driving scenarios. The main novel ingredient is the adaptation of the proposed driver safety platform operation to the individual driver behavior (e.g., aggressive driving) and his/her current biological state (e.g., attention level). We have developed a robust driver monitoring platform consisting of automotive sensors (i.e. OBD-II) that capture the real-time information of the vehicle and driving behavior as well as a heterogeneous wearable body sensor network that collects the driver biometrics (e.g., electroencephalography (EEG) and electrocardiogram (ECG)). In this investigation, we intend to examine the effect of the driving condition on the driver distraction as one aspect of the driver monitoring platform. Distraction during driving has been identified as a leading cause of car accidents. Our aim is to investigate EEG-based brain biometric measures in response to driving distraction. Using our proposed driver monitoring platform, we study driver cognition under real driving task in two different road conditions including of peak and non-peak traffic periods. Five subjects are recruited in our study and their EEG signals are recorded throughout the driving experience. The experimental results illustrated that the power of theta and beta bands in the frontal cortex were substantially correlated with the road condition. Our investigations suggested that the features extracted from the time-frequency brain dynamics can be employed as statistical measures of the biometric indexes for early detection of driver distraction in real driving scenarios.