BRAIN-IoT: Model-Based Framework for Dependable Sensing and Actuation in Intelligent Decentralized IoT Systems

Modern applications in the Smart Building and Industry 4.0 scenarios will be complex software ecosystems with strict requirements of geographic distribution, heterogeneity, dynamic evolution, security and privacy protection, highly more challenging than the ones required by the current environments. Two of the main challenges arising in the current Internet Of Things scenarios, i.e., the Smart Building one, are, on one side, the requirement of interconnecting several heterogeneous platforms and smart Things in the same environment and, on the other side, the need to be able to evolve the complex software ecosystem deployed, reacting automatically and at runtime to environmental changes, without the human intervention. To address these challenges, BRAIN-IoT establishes a framework and methodology supporting smart cooperative behaviour in fully de-centralized, composable and dynamic federations of heterogeneous Internet of Things platforms. In this way, BRAIN-IoT enables smart autonomous behaviour in Internet of Things scenarios, involving heterogeneous sensors and actuators autonomously cooperating to execute complex, dynamic tasks. Furthermore, BRAIN-IoT enables dynamically deploying and orchestrating distributed applications, allowing the automatic installation and replacement of smart behaviours reacting to environmental changes and User events. Finally, BRAIN-IoT provides a set of components that guarantee the security and privacy protection of the data exchanged using the solution. BRAIN-IoT is a general purpose solution that aims at being adaptable for heterogeneous scenarios, from Service Robotics to Critical Infrastructure Management. This paper introduces a Smart Building use case of the solution, which allows highlighting the advantages given by BRAIN-IoT in such scenario.