Artificial Intelligence (AI) is potentially useful for cost effective diabetes self-management. One research priority for the development of robust and beneficial AI concerns the use of formal verification techniques to model such self-modifying systems. In the context of diabetes, formal methods may also have a role in fostering trust in the technology as well as facilitating dialogue between a multidisciplinary team to determine system requirements in a precise way. In this paper we show how the formal modelling language Event-B can be used to capture safety-critical constraints associated with AI systems for diabetes management.
Brown, DanielMartin, ClareDuce, DavidAldea, ArantzaHarrison, Rachel
Faculty of Technology, Design and Environment\Department of Computing and Communication Technologies
Year of publication: 2017Date of RADAR deposit: 2017-07-18