Fully autonomous vehicles are coming — but until autonomous systems are able to account for every situation, every environment and every circumstance, automated driving will continue to be a continuum, with humans an integral part of the driving team.
Progressing along that continuum presents distinct challenges. Today, many OEMs are moving to advanced partial automation at Level 2+ and conditional automation at Level 3, balancing performance and affordability by allowing the driver to disengage from certain driving tasks for periods of time. The flipside of that benefit is that there will be times when the system requires the human to reengage and take control of the vehicle.
Ensuring that the handoff is as seamless as possible should be the automated system’s goal as much as it is the human’s. What is required is an intelligent approach to interacting with the driver — one that combines an environmental model of both the inside and outside of the vehicle with a driver model built with contextual assistance and semantic understanding, to provide drivers with the tools they need to successfully accept the transition of control.
