Human Activity Inference for Ubiquitous RFID-Based Applications

Radio frequency identification (RFID) is an important enabling technology for context-awareness in ubiquitous computing. RFID technology with complementary sensors is widely considered as a very promising approach to realizing the vision of ambient assisted living (AAL), which is an emerging application class in the domain of ubiquitous healthcare services. At the core of any assisted living systems is the important function of human activity inference, which is what enables such systems to be intelligently perceptive and responsive to the humans under their care. In this paper, we review the current state-of-the-art in activity inference for RFID-based applications, and present our ongoing work on a RFID-based assisted living prototype for `goal training´ or brain rehabilitation of patients with cognitive impairment in their home environments, with a discussion on the potential design issues involved.