Empirical research on car-following and lane-changing: Recent developments, emerging vehicle technologies’ impact, and future research needs

• Reviewed car-following and lane-changing empirical research. • A two-stage review framework is presented. • Research needs in car-following and lane-changing, and its link to microscopic models are elicited. • Need of new models in the era of connected and automated vehicles is emphasised. • Rigorous efforts for data collection with human factors are required. A sound understanding of driving behaviour is indispensable for properly modelling and predicting traffic dynamics, and accurately evaluating its impact on efficiency, safety, energy consumption, and the environment. This understanding becomes more critical in the era of emerging vehicle technologies, which are expected to significantly influence driving behaviour. Thus, this paper reviews empirical research efforts dedicated to understanding driving behaviour. Particularly, longitudinal (car-following) and lateral (lane-changing) driving tasks for conventional, connected, and automated vehicles are reviewed. Car-following behaviour is reviewed using different characteristics, like response time, time headway, etc., whereas lane-changing behaviour is reviewed from three aspects: decision-making, execution, and impact. This paper envisions several future research directions by identifying the limitations of existing research efforts and linking them to microscopic modelling. Also, realising emerging vehicle technologies are on horizon, the current empirical research on driving behaviour related to these vehicles is thoroughly reviewed, and research needs are emphasised.