The expanded deployment of smart objects in IoT applications is pushing existing IoT platform architectures and their security functionalities to their limits. Indeed, smart objects exhibit semi-autonomous behaviours, are not centrally controlled all the time and therefore need more dynamic approaches in protecting them against vulnerabilities and security incidents. In this paper, we introduce a novel framework for securing the latest generation of IoT applications that involve smart objects, while illustrating its application in securing an Ambient Assisted Living (AAL) system that comprises socially assistive robots. The framework’s innovative aspects lie in the use of predictive analytics for anticipating the behaviour of smart objects, including abnormalities in their security behaviour. The importance of anticipating such abnormalities is validated, demonstrated, and discussed in the context of the AAL application.
Keywords: Internet of things; IoT; Security; Smart objects; Robots; AAL; Ambient assisted living; Socially assisted robots; Healthcare