Current Open Positions
PhD Position available at the School of Computer Science, University of Birmingham, UK
Title: Integrated Safe & Robust Decision Making in Evolutionary Networked Cyber-Physical Systems
Project description
Emerging applications of Cyber-Physical Systems (CPS), such as multi-agent robotics, Internet-of-Things, autonomous driving, and smart manufacturing often involve a large number of hardware and software components that interact with each other in a coupled, time-sensitive, and context-varying operation environment. Together they create a networked system with physical interconnections and distributed data exchange that is challenging to safely and robustly design, regulate, and control in an intelligent fashion. Main reasons for complexity of such systems pertain to the highly dynamic and heterogeneous characteristics of the networked components, time-varying and evolutionary pattern of interactions between them, and the lack of architectural integrity between cyber and physical elements. On the other hand, these networked systems are often safety-critical, hence, safe decision making in a robust regime is necessary to achieve for reliable performance satisfaction. Any deployable and realistic AI-driven design framework for Networked-CPS should systematically take the mentioned challenges into consideration in order to sufficiently respond to safety and robustness requirements of automated networked systems in a formal and verifiable way. Our research in this area will use AI, data-driven models, control methods, and verification tools to develop novel integrative design, interaction, regulation, and control methods to guarantee safe and robust decision making across a web of heterogeneous and dynamic components that together cooperate to achieve a global objective.
This position will address foundational challenges within one or more of the following focus areas:
Structural analysis of evolutionary networked systems and operation boundaries
Modular safety and adaptive decision making for autonomous networked systems under evolutionary network topology
Robustness guarantees with respect to model and data uncertainty/reliability
Smart coordination of actionable information between distributed components with required communication accuracy (reliability, latency, throughput, and privacy)
Computationally efficient optimal policies (model-/data-driven) with timeliness guarantees
Candidates requirements:
Proficiency in mathematics, especially in linear algebra, statistics, multivariable calculus, and optimization is required.
Familiarity with, or keen interest in, optimal control, stochastic processes, networked systems, and formal methods is desired.
Strong programming skills, preferably in Python/C++, and Matlab.
Excellent oral and written communication skills.
Strong ability and interest in complex problem-solving and commitment to excellent research.
Experiences of presenting or preparing scientific manuscripts in journals or conferences is nice to have·
For more details about this position and application, please contact at: m.h.mamduhi@bham.ac.uk