Welcome to my site where you can find collected information about my academic and commercial work.
My research explores a novel synthesis of software engineering, model-based design, and AI techniques for the design and operation of dependable systems. Over the past 20 years, I have pioneered a body of work on model-based dependability analysis and evolutionary optimisation of complex engineering systems known as Hierarchically Performed Hazard Origin and Propagation Studies (HiP-HOPS). I have also co-authored EAST-ADL, an emerging architecture description language.
I am working in the area of Dependable AI and with my research group (DEIS-RG) innovate on topics including a) new metaheuristics for natural computation b) safety of machine learning c) dynamic safety assurance of cyber-physical and autonomous systems and systems of systems and, d) trustworthy and explainable AI. My research has significant economic and social impact with global reach, which is extensively documented in publically available Research Excellence Framework impact case studies BIOLOGIC, VETA and HiP-HOPS.
​
I am also researching digital art and various aspects of philosophy and its intersections with science. I believe in collaborative research environments that combine the intellectual pursuit of knowledge with positive outcomes for society, and develop the philosophically inspired concept of the research environment as "Academic Garden". Below you will find hyperlinks to a few of my current projects:
​​
-
HIP-HOPS tool site
​​
-
Simulation X + HiP-HOPS plugin with ESI-ITI GmbH (Germany)​
​​
-
EAST-ADL module in Metaedit+ with capabilities for safety analysis by Metacase (Finland)
​​
-
Dependable AI (see among other things a virtual urn with a prelaphaelite image of Pandora created in TIMAEUS, linking Nick Bostrom's "vulnerable world hypothesis" to the work of my group in this area)
​
-
SESAME (Safety and Security of Multi-Robot Systems) H2020 Research Project and DEIS (Dependability Engineering Innovation for Cyber-Physical H2020 Research Projects (2017-2024) developing the concept of Executable Digital Dependability Identities (EDDIs), for dynamic safety assurance in complex Systems of Systems such as autonomous cars and multi-robot systems
​
-
SafeML (Safety of Machine Learning) - Project with Fraunhofer IESE and Nuremberg Institute of Technology developing statistical techniques for making Machine Learning safer via the detection of distributional shifts between operational and training data
​
-
EPSRC EDGE AI Hub Large EPSRC project investigating Edge-AI. Our role is to improve the safety of this technology by applying SafeML and EDDIs to ensure that faults, inaccuracies, and threats are dealt with appropriately on intelligent systems on the Edge.
​
-
DREAM (Data-driven Reliability-centred Evolutionary Automated Maintenance for Offshore Wind Farms - A project funded by EDF-London
​
-
Philosophy and digital art projects: TIMAEUS, VIRTUAL STOA,​ GeNeRaTiVe aRt project, ODYSSEY
