Design of Indoor Occupancy Estimation System with Wireless Brain-Inspired Computing Platform

The estimation of indoor occupancy rates has become increasingly important, especially for accident prevention at events and providing critical rescue information during disasters. However, conventional systems require significant effort and cost for installation, such as the need for dedicated servers and internet connectivity, making them difficult to deploy conveniently. This study proposes the design of an indoor occupancy estimation system based on the Wireless Brain-Inspired Computing (WiBIC) platform. By equipping each IoT device with neuron functionality, environmental data is compressed into spike signals directly at the source. These signals are processed using intelligent information processing techniques based on the principles of brain-inspired computing, enabling the estimation of indoor occupancy. The WiBIC platform's serverless architecture and low-power characteristics allow for easy and efficient deployment of the system in various indoor environments. The proposed system further emphasizes energy efficiency by incorporating FPGA implementation and the Asynchronous Pulse Code Multiple Access (APCMA) protocol. Experimental evaluations were conducted using data collected from illuminance sensors, human detection sensor, and current sensors to estimate occupancy rates. The results show that the system achieves practical accuracy, validating its effectiveness. The accuracy achieved is sufficient for practical applications, demonstrating the effectiveness of the proposed system.