When edge intelligence meets cognitive buildings: The COGITO platform

Future buildings are complex systems that aim at improving the quality of life of their inhabitants and increasing safeness, security, and efficiency. In order to reach these goals, they require their own self-management and self-adaptation capabilities, thus becoming cognitive entities. However, developing cognitive buildings that exploit advanced Artificial Intelligence (AI) techniques in a distributed fashion is still a challenge. Indeed, they need to continuously collect and process a variety of environmental parameters, learn and predict the users' needs and preferences, and then control a large number of heterogeneous devices. These operations may leverage the dynamic availability of edge and cloud computing resources. This paper discusses the challenges and requirements tied to the development of cognitive buildings and proposes COGITO, an Internet of Things platform specifically suited to design and implement cognitive buildings. COGITO enables edge intelligence and the exploitation of all the available resources in the computing continuum between the edge and the cloud. All of this is to support holistic services that guarantee energy efficiency, security, and user comfort. For validation purposes, a case study involving the use of COGITO in an office building is presented, together with a performance evaluation analysis.