Given the long lifetime and high cost of the built environment, it is imperative that we plan for and create communities that are robust in the face of climate change. Occupancy detection systems are widely used to monitor and detect events for building management purposes. Effective adaptation in the built environment needs to be supported by robust policy and a range of incentives to ensure delivery on the ground and likewise the need for artificial intelligence (AI) to help in these intervention is of utmost importance. This study proposed the use of an indoor occupancy detection using ultrasonic and Carbon Dioxide Sensors to help determine the usability of energy by occupants in a buildings space. Current research has presented solutions to improve occupancy detection and estimation to enhance building performance. The majority of these solutions use occupant static or dynamic data and indoor metrological data to generate automated occupancy detection and estimation models to manage HVAC operation to ensure significant energy saving through programmable microcontroller via wi-fi or mobile application. These solutions have been proven to be economical through improved overall HVAC energy consumption efficiency compare to a traditional thermostat. However, recent investigation shows these solutions suffer from false human occupancy detection and poor estimation strategy. Incorrect detection was caused by technological hardware and research limitations. The estimation strategies are based on estimation and approximation without acknowledging the base of the assumption and resulting in excess energy consumption, especially when space is occupied by less designated occupancy. This study proposed a sensor fusion mechanism to accurately collect and analyze occupancy information to accurately classify occupancy as a space presence. The experimental analysis shows 89% accuracy of human occupancy detection without compromising occupancy privacy