I always look for motivated PhD students and postdoctoral scholars. Potential candidates should have solid background in (at least) one of the following areas: Systems and Control, Mathematics, Computer Science, Artificial Intelligence, Data Science, Operations Research, or a closely related field. If you are interested, please send your application to abolfazl.lavaei@newcastle.ac.uk. If you possess outstanding programming skills in Python/C++ along with an M.Sc. degree in CS/ML/Robotics, and aspire to pursue your PhD studies in theoretical and practical aspects of software development for cyber-physical systems, please send your CV detailing programming experiences to abolfazl.lavaei@newcastle.ac.uk with the subject line “PhD: CPS - Software Development”. Current openings
Title: Towards Safe and Efficient AI-enabled Control of Autonomous Vehicles via Quantum Computing Overview: The emergence of autonomous vehicles (AVs) brings immense potential benefits including the ability to save human lives by reducing the number of traffic collisions, and achieve Zero Carbon Futures by improving energy efficiency and alleviating traffic congestion. While the technological strides in AVs offer numerous privileges, the efficient provision of safety certification has consistently posed a significant obstacle to their successful deployment. This project aims to leverage the power of quantum computing to enhance the safety and efficiency of AI-enabled control mechanisms for autonomous vehicles. The proposed research is critical for addressing the challenges associated with real-time decision-making in complex and dynamic environments. This involves interdisciplinary research combining disciplines of formal methods, control theory, and data science. If you are interested, please send your application (CV, transcripts, etc.) to abolfazl.lavaei@newcastle.ac.uk, by February 2024, with the subject line “PhD: Quantum Computing in AVs”.
Title: Safe AI-enabled Control of Autonomous Vehicles Overview: Autonomy is certainly one of the main themes of the 21st-century technology. In the near future, we expect to see fully autonomous vehicles (AVs), aircrafts, and robots, all of which should be able to make their own decisions without direct human involvement. Although such technological advances for AVs bring many potential advantages, e.g., fewer traffic collisions, reduced traffic congestion, increased roadway capacity, relief of vehicle occupants from driving, etc., providing safety certification and guaranteeing correctness of the design of such autonomous vehicles have been always the major obstacles in their successful deployment. Over the past decade, several advances have been made in developing various techniques for AVs, especially in the areas of perception, planning and control. However, a stumbling block in their deployment is the lack of any guarantee of correctness. This in turn opens several unanswered questions related to liability in the case of accidents or other unforeseen events. The main goal of this interdisciplinary project is to design a safe AI-enabled controller framework for autonomous vehicles by requiring knowledge from control theory, data-driven methods, and SMT solvers. International collaborations will be expected with experts in safety-critical systems together with industrial collaborations. If you are interested, please send me your application (CV, transcripts, etc.) by February 24.
Title: Towards Trustworthy AI for Safety-Critical Cyber-Physical Systems Overview: Cyber-physical systems (CPS) are complex networked models combining both cyber (computation and communication) and physical components, which tightly interact with each other in a feedback loop. Examples of such systems span a wide range of real-life safety-critical applications including automotive, robotics, transportation systems, energy, healthcare, critical infrastructures, and so on. Formal verification and policy synthesis for this type of complex systems to enforce high-level control missions are inherently very challenging mainly due to (i) large dimension of underlying systems, (ii) stochastic nature of dynamics, (iii) tight interaction between physical and cyber components, (iv) dealing with complex requirements, and (v) lack of mathematical closed-form models in many real-world applications. The main goal of this research is to focus on developing scalable AI-based techniques with mathematical guarantees to tackle the aforementioned difficulties and design highly-reliable CPS by bringing together interdisciplinary concepts from formal methods in computer science, optimization in operations research, control theory, data science and machine learning.
|