ESR 2 – Self-learning for Optimum Manufacturing Equipment (Individual & Collective Response)
It is recognised that current adaptive control systems cannot effectively cope with unanticipated external or internal disturbances. The project will focus on developing machine learning methods, building upon existing works such as reinforcement learning, to allow both individual and collective self-learning and self-certification within a regulated environment. The learning process will utilise the status data, past experiences and key parameters to analyse possible scenarios and propose actions for achieving the individual and collective goals. Newly learned skills, combined with the context in which they are relevant, will supplement the predefined set of user-supplied skills and will be used individually or collectively to respond promptly when similar situations arise in the future.
University of Nottingham (UNOTT), United Kingdom
UNINOVA, KTH, STIIMA, UCTM
UNOTT - Prof Svetan Ratchev, Prof Atanas Popov, Dr Giovanna Martinez Arellano