Optimizing HVAC Systems for Sustainable Lecture Rooms: Harnessing Environmental and Occupancy Data for Comfort and Energy Efficiency through Data-Driven Insights

Abstract

The escalating energy consumption in campus infrastructure, especially in lecture halls with heating, ventilation, and air conditioning (HVAC) systems, necessitates data-driven optimization strategies. This research demonstrates the integration of Internet of Things (IoT) sensors with cloud-based predictive analytics to develop intelligent lecture room policies aimed at enhancing efficiency and sustainability. A Raspberry Pi-based IoT device, equipped with a BME680 sensor for monitoring temperature, humidity, and air quality, and a passive infrared sensor for occupancy detection, was installed in a university lecture room for real-time data acquisition. Data collected was routed through MySQL for storage and Node-RED for preprocessing. Time series forecasting models, including ARIMA and Prophet, along with machine learning models like XGBoost, achieved over 90% forecast accuracy for temperature and occupancy levels, enabling proactive control of environmental conditions. The optimized HVAC scheduling, based on forecasted occupancy patterns, resulted in a 20% reduction in energy consumption over an 8-week deployment, ensuring thermal comfort by maintaining temperatures within the recommended range of 21-23°C during occupancy. Enhanced occupant comfort was also achieved by maintaining humidity levels between 40-60%, improving indoor air quality through proactive ventilation control. Key recommendations include dynamic HVAC scheduling based on occupancy forecasts, thermostat setpoint adjustments to prevent temperature peaks, and expanding IoT sensor deployments across campus facilities to generate deeper insights. This integrated IoT and predictive analytics approach enabled a sustainable and responsive built environment, providing a scalable framework for optimizing other infrastructure types such as laboratories and offices.