UKRI Future Leaders Fellowship
“Globally, 1.25 million people die on the roads due to road accidents, and it’s said that 90% of these accidents are due to driver error. So if we could assist the driver or remove the driver, there is a way to reduce this fatality number.”
-- FELLOWSHIP OBJECTIVES --
To identify test scenarios and test cases for evaluating CAV to identify which are the smart miles (i.e., situations where CAV fail) using various analysis methods including machine learning (ML) based methods like Bayesian Optimisation, SOTIF (Safety Of The Intended Functionality) analysis, and real-world data annotation
To develop a National CAV Test scenario catalogue that will help coordinate research activity with various collaborative R&D projects
To create a Scenario Description Language for a common understanding of scenarios
To generate knowledge to enable test scenario selection for test environments (digital world, simulated environment, test tracks and real-world)
To generate knowledge of how to claim safety for ML-based systems
To develop a methodology for safety argumentation of CAV using simulation-based and real-world based safety evidence
To create requirements and validation methods for simulation tools
Disseminate research findings to society
Inform the standards, policy, and legislative frameworks for the UK on testing, evaluation, and deployment of autonomous vehicles
Incorporate the learning from the research into all levels of engineering education
-- The Scheme --
The inaugural UKRI Future Leaders Fellowship Scheme will receive £900 million over the next 11 years, with six funding competitions and 550 fellows set to be awarded. It will help to establish the careers of world-class research and innovation leaders across UK business and academia, providing a boost to the pipeline of talent needed for the future. The scheme supports early career researchers and innovators with outstanding potential in universities, UK registered businesses and research-focused environments such as research councils' institutes and laboratories. The support will enable each fellow to tackle ambitious and challenging research.
Academic or business hosts will benefit from funding of up to £1.2 million over a four-year period. The programme will also provide hosts with an opportunity to attract and retain the most talented individuals (UK or international) to their organisation.
While prototype CAV technologies have existed for some time now, ensuring the safety level of these technologies has proven to be a hindrance to commercialisation of CAV. It is suggested that, in order to prove that CAV are safer than human drivers, they will need to be driven for more than 11 billion miles, an unfeasible proposition. Furthermore, industry trends in CAV suggest the widespread adoption of Machine Learning (ML) in the autonomous control systems. ML-systems are non-deterministic in nature, resulting in different behaviours and a lack of transparency around the CAV system. Therefore, it is often difficult to identify reasons for a particular failure in such ML-based systems and take the corrective measures.
Siddartha's UKRI Future Leaders Fellowship (FLF) programme will develop pioneering testing standards to enable robust and safe use of CAV with a focus on creating fundamental knowledge as well as research methods and tools. With safety as an underlying theme, Siddartha's FLF will focus on testing scenarios and environments using simulation, and creating evidence to prove CAV and ML-based systems are safe.
As the UK's representative on various international standards committees, Siddartha will aim to help position the UK as a world leader in CAV research and innovation, for a long lasting societal and economic benefit, whilst also creating a world leading team of researchers in CAV evaluation at WMG. The programme's research outputs will provide a clear route to deliver impact through the development of international standards.