Occupant behaviour is one of the key parameters that significantly impact the operation of heating, ventilation, and air-conditioning (HVAC) systems and the energy performance of buildings. The detailed occupancy information can improve HVAC operation and utilisation of building spaces. Strategies such as vision-based occupancy detection and recognition have recently garnered much interest. This study investigates the performance of a vision-based deep learning detection technique for enhancing building system operations and energy performances. The model used was the Faster RCNN with Inception V2. Two occupancy detection model configurations were developed, tested and evaluated. Both models were analysed based on the application of the detector within a selected case study building, along with the evaluation based on the different evaluation metrics. Results suggest that the occupancy detector (Model 1) provided an overall accuracy of 95.23% and an F1 score of 0.9756, while the occupancy activity detector (Model 2) provided an accuracy of 89.37% with an F1 score of 0.8298. Building Energy Simulation (BES) was used to evaluate and compare the impact of such an approach on the indoor occupancy heat gains. The study highlighted the potential of the detection approaches, but further development is necessary, including optimisation of the model, full integration with HVAC controls and further model training and field testing.
Tien, P. W., Wei, S., Chow, T. W., Darkwa, J., Wood, C., & Calautit, J. K. (2022). Enhancing the detection performance of a vision-based occupancy detector for buildings. Proceedings of the ICE - Engineering Sustainability, https://doi.org/10.1680/jensu.22.00013