Using Brain-Computer Interface in Cognitive Buildings: a Real-Time Case Study

Cognitive Buildings (CBs) are IoT-based smart infrastructures that self-learn, self-organize, and adapt through Artificial Intelligence (AI), thereby enhancing occupant satisfaction with minimal human intervention. Recent advances in wearable technology have enabled the integration of electroencephalogram (EEG) sensors into CBs, allowing continuous, unobtrusive monitoring of residents' states. In this paper, we propose a novel Cloud-Edge Brain-Computer Interface (BCI) system that leverages Steady-State Visual Evoked Potentials (SSVEPs) to provide real-time feedback to CBs. Our approach is validated through a comprehensive case study conducted at the IoT Laboratory of the ICAR-CNR headquarters in Rende (Italy), where a consumergrade wearable EEG headset is integrated with embedded, edge, and cloud devices. Furthermore, we evaluate the performance of four machine learning algorithms deployed on an embedded Raspberry Pi 4 for SSVEP recognition, demonstrating the system's potential for real-time applications in CB environments.