About the Project
The successful candidate will perform research to improve and validate the performance of autonomous vehicle (AV) sensor perception, specifically RADAR image recognition, to improve overall AV risk management and real-time decision making. The safety assurance of AVs in land, ocean, and aerial modes of operation is not yet validated. Shortcomings with onboard sensor perception render AVs vulnerable to losing situational awareness in safety critical situations, close to hazards, and in unsegmented traffic.
Miniaturised RADAR has recently become commercially available with weight, size, and energy consumption profiles suitable for deployment on small AVs, extending detect-and-track of non-cooperative obstacles to longer ranges (c. 2km) and poor visibility conditions (night time, rain / fog). However, issues with signal noise and multipath in the dynamic and complex theatres characteristic of AV use cases present outstanding safety challenges. This PhD project will extend the performance of AV RADAR software to cluttered scenes involving urban industrial transport systems and infrastructure. You will join a team of researchers developing massively parallel simulations and artificial intelligence to deep mine resilient context awareness into AV navigation, guidance, and control. The validation use case directly benefits sectors such as transport, construction, energy, and security. The outcomes will inform the development of emerging and evolving regulatory frameworks and unlock high value AV services requiring operation in close proximity to hazards (such as maintenance-inspection, last mile delivery, and autonomous shipping).
The successful candidate will join an excellent team of researchers at the ALMADA Research Centre in the new flagship UWS campus in Lanarkshire. The ALMADA team is currently leading a £5 million research portfolio into autonomous vehicles, including the European Commission H2020 project RAPID. Working within the ALMADA research environment, the successful candidate will engage with an international partner network, avail of field testing opportunities, have access to specialised drone equipment and HPC hardware, and benefit from the advice and mentoring of international experts.
The successful candidate will spend extended periods at Codeplay Software. The company is headquartered in Edinburgh, with over 70 employees, and is a leader in software portability and the development of software tools and supporting technologies for heterogeneous multi-core processor systems. Building on its proven expertise in games, mobile phone and HPC software, Codeplay has developed open standards and new products that address the pain points of deploying Artificial Intelligence applications into new HPC platforms and mass-produced devices, with a focus on the automotive sector, specifically Advanced Driver Assistance Systems (ADAS).
Eligibility & entry requirements:
Entry into this exciting research programme is competitive and will take account of qualifications, aspirations and experience. Successful applicants are likely to have a computer science background ideally with some specialisation in one or more of visualisation, artificial intelligence, GPGPU, and software engineering; students from related fields will also be considered. Students are normally expected to have a 2:1 or 1st class honours degree from a UK University or an equivalent standard from an overseas university. The successful applicant must be eligible for UK and EU fees and the funding currently offers stipends of £16,000 plus fees, training, travel, and consumables.
UWS supervision team - Dr. James Riordan and Dr. Paul Keir ([Email Address Removed], [Email Address Removed]).
Chief Scientist at Codeplay - Dr. Jens-Uwe Dolinsky ([Email Address Removed]).
Keywords: “modelling and simulation”, “artificial intelligence”, “sensors”, “autonomous vehicles”, and “computer vision”.
Based on your current searches we recommend the following search filters.
Based on your current search criteria we thought you might be interested in these.