Occupancy tracking in indoor environment

Abstract

Smart buildings incorporate advanced sensor technology and IoT systems to provide people with a more comfortable indoor living experience. So, accurately estimating indoor occupancy is critical to further optimizing building energy use. However, environmental factors and privacy concerns often challenge traditional sensor-based approaches. To overcome these challenges, this paper proposes an innovative approach to tracking people indoors using Wi Fi technology and AI algorithm technology without relying on traditional sensors. It only needs to be based on Wi Fi signal strength for non-intrusive people monitoring. By combining artificial intelligence technology and fingerprint databases, this approach improves the accuracy of predicting indoor occupancy. During the development process, the WiFi signal strength (RSSI value) must be captured to build a fingerprint database. Then, the KNN and CNN algorithms were used to make predictions on the input. Finally, the outputs of the two models were integra ted to im prove the accuracy of occupancy tracking further. This method ensures that the difference between the predicted and accurate coordinates does not exceed a radius of 0.6 meters. This innovative approach provides a more efficient and intelligent solution for the prediction of indoor occupancy while also protecting users' privacy and providing them with a more comfortable indoor living experience.