Partners: Wayve Technologies Limited (Lead), University of Warwick.
DriveSafeAI is a project between Wayve and WMG at the University of Warwick, and will develop AI assurance methodologies to ensure the safe deployment of self-driving vehicles, paving the way to commercialisation of this technology globally.
Demonstrating the safety of self-driving systems is a crucial step towards realising the benefits of this technology. DriveSafeAI provides a solution, bringing together Wayve’s expertise in developing end-to-end machine learning systems and combining this with WMG’s world-leading expertise in verifying and validating the safety of self-driving technologies.
Wayve is reimagining self-driving vehicle technology, using data and machine learning to create an embodied artificial intelligence, with the capability to drive any electric vehicle in any city. Wayve’s unique approach, called AV2.0, can seamlessly integrate with existing fleets and infrastructure and adapt to new environments.
DriveSafeAI will develop a set of safety methods, tools and datasets for self-driving vehicles, and an Independent Advisory Committee will provide expert input from across the self-driving supply chain to ensure the project builds confidence and trust in AV2.0.
DriveSafeAI will develop an AI assurance methodology capable of scaling to different Operational Design Domains (ODDs).
The project will develop a novel OASISS (ODD-based AI Safety In Self-Driving Systems) approach which can be generalised and applied to new ODDs on operational timescales, ensuring facile verification and validation of AI-based software for new geographies .
The methodologies developed by DriveSafeAI will support the whole self-driving industry, for systems which have discrete AI components, rather than just end-to-end machine learning used in Wayve’s AV2.0.
Finally, the project will leverage Wayve’s simulation suite to provide realistic and diverse simulation scenarios to train self-driving vehicles enabling the vehicles to experience edge-case scenarios, many of which would not be possible to test in the real world. DriveSafeAI will develop an ODD and behaviour-based scenario training and testing approach for the Wayve Infinity Simulator to realise complex and challenging driving scenarios that allow Wayve to train, understand and validate AV2.0’s driving intelligence.
Aims to develop an AV capable of safely driving in residential, urban, and rural environments.
A collaborative initiative to create an affordable, robust navigation system for automated vehicles.
Provides a toolset that helps to efficiently identify, define and execute the test requirements for an ADS.
Aims to develop advanced position and navigation sensors that work reliably in various environments.
Aims to enable accurate representation of ADS sensors in simulation.
Focuses on creating an autonomous dolly for airside cargo movements.
Aims to deliver a universal and affordable drive-by-wire system that replaces traditional mechanical linkages with electronic ones.
Aims to develop a high-performance imaging radar product specifically designed for AVs.
DeepSafe aims to support the verification and validation (V&V) of automated driving systems (ADS).
A ‘plug-and-play’ roadside connectivity solution.
A fully redundant, fail-operational Drive-by-Wire technology platform to enable safe driver-out, on-road autonomous vehicle capability.
Focuses on the development of a modular dual redundant steer-by-wire system for heavily automated and electric vehicles.
To learn more about the CAM Supply Chain UK competition and the remarkable projects that have been awarded, contact competitionsupport@zenzic.io
{{ teamMember.name }}
{{ teamMember.title }}