AHSWN Home · Issue Contents · Forthcoming Papers

Comprehensive Review of Sensor Data Fusion Techniques in IoT-Based Occupancy Detection Systems
Pushpanjali Kumari, S R N Reddy and Richa Yadav

In smart environments, occupancy detection is critical in maximising energy efficiency and enhancing the consumer’s experience, particularly in sensor-based Internet of Things (IoT) platforms. By applying machine learning (ML), deep learning (DL), and transfer learning (TL) approaches to occupancy detection through IoT sensor fusions, the study advances the subject. The study also investigates current occupancy detection methodologies, focusing on integrating data from motion detectors, temperature sensors, and environmental monitors. Despite these systems positively impacting building management accuracy and responsiveness, challenges in security and privacy persist. The paper explores recent developments, including predictive occupancy models and privacy-preserving detection systems. However, the limited availability of comprehensive frameworks addressing security and privacy concerns highlights the need for innovative solutions. This paper emphasises that addressing these challenges and promoting interdisciplinary collaboration will propel the evolution of occupancy detection towards more secure, privacy-aware, and user-centric systems in the dynamic landscape of smart buildings.

Keywords: Internet of Things, occupancy detection, security, privacy, machine learning, deep learning, transfer learning