Emerging complex smart environments are characterized by the presence of a number of interconnected sensors and devices, forming the so-called Internet of Things. In such environments various activities by human, digital or robotic actors take place. Research on Activity Recognition (AR) has typically followed two complementary directions: data-driven approaches, exploiting machine learning techniques, and knowledge-driven approaches, leveraging logical modelling and reasoning. This talk focuses on the latter category, presenting a rule-based approach for both offline and real-time recognition of so-called Activities of Daily Living (ADL), relying on events produced by a non-intrusive multi-modal sensor infrastructure deployed in a residential environment, as part of the SPHERE project. Novel aspects of the approach include: the ability to recognise arbitrary scenarios of complex activities using bottom-up multi-level reasoning, starting from sensor events at the lowest level; an effective heuristics-based method for distinguishing between actual and ghost images in video data; and a highly accurate indoor localisation approach that fuses different sources of location information. The proposed approach is implemented as a rule-based system using the Jess rule engine and is evaluated using data collected in a smart home environment of the SPHERE project. Experimental results show high levels of accuracy and performance, proving the effectiveness of the approach in real world setups.
Dr. George Baryannis received his Dipl.Eng. in Electronic and Computer Engineering from the Technical University of Crete, Greece and his M.Sc. and Ph.D. in Computer Science from the University of Crete, Greece. He is currently a postdoctoral research assistant at the University of Huddersfield in the United Kingdom. His research interests include: Service-Oriented Computing; Semantic Web and knowledge representation and reasoning; Internet of Things; Cloud Computing. He is an IEEE and IEEE Computer Society Member.