Enabling a Novel Evaluation Continuum for Connected & Autonomous Vehicles

点击翻译按钮获取中文摘要

Moving people and goods is worth over £100 billion to the UK economy (across transport modes), but it comes at a cost, with over 2000 deaths and 120,000 injuries every year. Connected and autonomous transport has the potential to make land, air, and marine journeys safer, faster, and more efficient, contributing to both our national health and carbon emissions goals. Additionally, the connected and autonomous transport systems' market globally is projected to be over £700 billion by 2030, so this sector could be a major driver of economic growth in the UK. The biggest challenge to delivering the potential of autonomous transport systems are provable quantified safety and consumer understanding. Without addressing these issues across all sectors, it will take us significantly longer to unlock the potential commercial and wider benefits.Over the last three years of the Future Leaders Fellowship (FLF), to find answers to various research questions in Connected and Autonomous Vehicles (CAVs) (i.e., land), the fellow has often looked to other transport domains like aviation and marine and benefitted by transferring learnings from them to CAV. This experience brought a realisation that "while aviation and marine are also introducing the autonomous system, the safety challenges are similar to CAVs". This realisation underpins the fellow's +3 FLF renewal vision.The vision for +3 Renewal of the UKRI FLF application is to translate the learnings on safety assurance of CAV to aviation and marine autonomous systems (aerial drones and unmanned vessels). While there are obvious differences between the transport domains (land, air and marine), the approach to safety assurance could potentially be similar if a first principles approach is taken. Safety assurance of autonomous transport systems in air and marine requires three key areas of research, standards, and regulation: 1) test scenarios; 2) test environment; and 3) safety argument. However, the safety assurance process needs to be underpinned with the correct and complete set of requirements definition for the system. As part of research on CAV, defining the Operational Design Domain (ODD) and behaviour capabilities of the autonomous system is fundamental to the requirements definition.An ODD (i.e., well-defined safe operating boundaries) is fundamental to any safety assurance process for autonomous systems. Operating conditions for the land include attributes like road type, weather type, type of actors (emergency vehicles, pedestrians), etc. (as per BSI PAS 1883 - fellow as technical author). While the operating conditions for aviation and marine will differ, the concept of ODD is transferable. Operating conditions for the sea could include attributes like current strength, wind speed and direction, salinity, water depth, etc. Operating conditions for air could include attributes like wind speed and direction, air density, fog, etc. However, a standard taxonomy concept still evades the industry, which will be one of the focuses of my +3yrs FLF.Another key aspect of safety assurance of autonomous systems in transport is the definition of safe behaviour. As a further part of the +3yrs FLF, the focus would be to create safe behaviour definitions by codifying the Rules of the Air and Rules of the Sea. Currently, Air Traffic Management (ATM) and Convention on the International Regulations for Preventing Collisions at Sea (COLREGs) define the rules of air and sea for human-driven vehicles. We will take an ODD and behaviour-based approach to codify the rules using first-order logic.In addition, the +3yrs FLF will also benefit from the fellow's first-hand experience as the UK's technical representative on various international standards committees, providing further insight and a clear route to deliver impact from the proposed research through the development of international standards and regulations, while also ensuring that the UK becomes a global leader in this area.