This thesis investigates a range of sensor-based methodologies for the detection of occupancy in indoor environments, with the objective of improving HVAC control and energy efficiency. The research was conducted in the Bäckström Conference Room at KTH, utilizing existing environmental sensors, including CO₂, temperature, humidity, a PIR motion sensor, and custom-designed Bluetooth Low Energy tracking systems developed on ESP32 microcontrollers.

Among the techniques assessed, CO₂ concentration emerged as the most reliable single indicator of occupancy, demonstrating a robust correlation (R² = 0.98) with the actual headcount. While temperature and relative humidity provided useful insights into the dynamics and stratification of the indoor environment, their individual predictive capabilities were found to be limited. The PIR sensor predominantly functioned to verify presence due to its less precise measurement resolution. The BLE-based approaches comprised two distinct methodologies. The first was a BLE device counter that passively detects nearby broadcasting devices, showing a strong correlation with occupancy data. A linear regression model developed using the device count achieved an R² value of 0.874 with a dataset characterized by a limited number of occupants and devices. The second methodology focused on analyzing the BLE signal strength (RSSI) between a transmitter-receiver pair. Although this approach did not facilitate occupant counting, it enabled the detection of occupant positioning and movement patterns within the space by examining signal attenuation characteristics.

Occupancy data were employed to develop adaptive occupancy schedules, which were subsequently implemented in IDA ICE building simulations. The objective of these simulations was to evaluate the potential energy savings achievable through occupancy-based HVAC control by comparing dynamic, sensor-informed occupancy profiles with conventional static profiles grounded in EN 16798-1 standards. Importantly, this research focused on assessing two distinct occupancy scenarios within the same HVAC system configuration, rather than analyzing different energy control technologies. The results indicated a significant reduction in annual HVAC energy consumption when authentic occupancy data were utilized, highlighting the critical impact of internal gains on energy demand. The Bäckström Room, characterized as a thermally stable and controlled environment, relies predominantly on the Mechanical Supply Air system for temperature regulation, with minimal contributions from active heating and cooling systems.