Situational Awareness to Shift between Autonomy levels in Automated Vehicles

Project ID: 2228cd1306 (You will need this ID for your application)

Research Theme: Artificial Intelligence and Robotics

UCL Lead department: Mechanical Engineering

Project Summary:

Background: We are currently experiencing a paradigm shift towards highly automated vehicles. Over the few next years, drivers will be provided with increasingly sophisticated autonomy features. Each level of autonomy requires a different level of the driver’s cognitive/physical intervention. During autonomy level transitions, the vehicle may require an engaged driver to monitor the system and assume control under conditions when the car cannot drive itself.

Critical gap: Transitions between different levels of autonomy cause safety risks when taking back handling control from the car evident by stats from the California Department of Motor Vehicles reporting that 34% of collisions between 2014 and 2020 were during the transition modes.

Aim/Objectives: The aim is to understand how to determine and manage the level of Situational Awareness (SA) that is required for shifting control between the driver and vehicle in highly automated cars. The objectives are:

1) to identify the driver’s SA, which can be linked to autonomy levels, by means of multi-model sensory data fusion through fusing and classifying physiological and behavioural sensors (brain activity/EEG and eye tracking glasses). 2) to optimise multi-modal feedback mechanisms (auditory, visual, haptic) to increase driver’s SA. 3) to provide directional haptic cues to guide safe and timely responses in critical emergency driving situations.

This interdisciplinary project will train the student in designing experiments in virtual environments (Unity) with human participants of diverse backgrounds, ethics, data management, Responsible Research & Innovation, data analytics, AI/Machine Learning (Tensor Flow/Python).

The student will partially work at IM@UCL, funded by EPSRC ECR equipment grant EP/S01800X/1), a driving simulator for highly automated cars located at UCL PEARL. Ansible Motion (the company who supported IM@UCL) agreed on a secondment for the PhD students and cover costs for software packages (in-kind contribution of £6k pa).