Abstract Occupancy attracts an increasing attention in recent building energy efficiency research through enabling more sophisticated control strategies. With the development of latest indoor positing systems (IPS), the facility managers are able to detect the geospatial distribution of building occupants. Based on such information, this paper proposes a demand-driven control system for HVAC control in large spaces to reconcile occupants’ thermal comfort demand and energy consumption. Comparing to conventional temperature and CO2 based HVAC control systems, the new approach integrates the indoor positions of occupants and a demand-oriented and PID-based ventilation control mechanism. An computational fluid dynamic simulation is constructed to valid the proposed control system. Air supply flow rate and temperature distribution are captured for three sample cases that have even and uneven occupancy distributions in the simulation. By avoiding overcooling and unnecessary cooling, the proposed approach could save a significant amount of electrical consumption from HVAC operation.