NeeMAS: A Need-Based Multi-agent Simulator of Human Behavior for Long-Term Drifts in Smart Environments

Early identification of long-term changes in the behaviour of people monitored with smart environment solutions is essential to prevent health decline. However, data collection and analysis of human behaviour are challenging and time-consuming. A potential solution consists in creating digital twins of the individuals to replicate the typical behaviours for advanced data analytics. The Assistive Technology Group (ATG) at Politecnico di Milano has developed NeeMAS (NEEd-based Multi-Agent Simulator), a novel simulator that effectively simulates human behaviour with physiological and social needs, cognitive decay, and behavioural drifts due to ageing or disease onset such as apathy or depression. NeeMAS simulates a senior care facility with several individuals, spending part of their time in their rooms and in part sharing common indoor and outdoor spaces interacting with other people. Experimental results show the feasibility and flexibility of the proposed approach for the generation of typical human behaviours and their drifts.