Control Architecture for Autonomous/Assisted Driving of EV

Number of Students : 1
Guides : Umanand L, T V Prabhakar

Hierarchical control issues that enable steering for fully automatic and/or assisted driving have many applications. The basal automatic steering should be based solely on vision sensors just like in human driving. All path estimation and collision avoidance need to handled primarily by this sensing mechanism. Robustness and safety improvements will need other sensors with hierarchical levels of protection and redundancy. One of the significant cause of accidents on highways is driver fatigue. Due to monotony of long drives, the driver goes in and out of sleep without realising it. Sleep detection in drivers and enabling assisted driving function seamlessly would go a long way in avoiding many fatigue related accidents. Developing the control strategy for autonomous steering and assisted steering is a potential challenge.
A spin-off application of the autonomous steering problem is platooning. Platooning is a term used for managing a fleet of trucks moving in a convoy. The trucks speed, direction and distances between them are managed by connectivity support and automated driving support systems. Platooning is not only useful for fleet management, but it can also be used for traffic management. An off-board master can control the on-board propulsion system at traffic junctions and certain pre-determined zones like school zones. This problem has many challenges in terms of not only autonomous steering and driving, but also inter-vehicle communication, collision avoidance, redundancy in masters etc.