Human-vehicle collaborative driving to improve transportation safety

Abstract

This dissertation proposes a collaborative driving framework which is based on the assessments of both internal and external risks involved in vehicle driving. The internal risk analysis includes driver drowsiness detection, driver distraction detection, and driver intention recognition which help us better understand the human driver's behavior. Steering wheel data and facial expression are used to detect the drowsiness. Images from a camera observing the driver are used to detect various types of driver distraction by using the deep learning approach. Hidden Markov Models (HMM) is implemented to recognize the driver's intention using the vehicle's laneposition, control and state data. For the external risk analysis, the co-pilot utilizes a Collision Avoidance System (CAS) to estimate the collision probability between the ego vehicle and other vehicles. Based on these two risk analyses, a novel collaborative driving scheme is proposed by fusing the control inputs from the human driver and the co-pilot to obtain the final control input for the vehicle under different circumstances. The proposed collaborative driving framework is validated in an Intelligent Transportation System (ITS) testbed which enables both autonomous and manual driving capabilities.