NTU Singapore and Japanese building automations company, Azbil Corporation, are collaborating to develop an integrated approach for improving thermal comfort and energy efficiency in office spaces through novel sensors and occupant feedback system under the EcoCampus initiative. The research project, which is first-of-its-kind, comprises two key processes to optimize the overall air-conditioning system.
One such process is the Thermopile Array Sensors (TAS), which is an intelligent positioning system (Figures 1 & 2) to automatically detect the location of occupants, plug loads and process loads in air-conditioned spaces. TAS is new generation of thermo-sensitive elements comprising infrared (IR) sensor arrays.
The second process includes Thermal Sensation Voting (TSV) installed at every occupant's workstation. The communication function enables occupants to vote for comfort feelings. Predicted Mean Votes from TSV (Figure 3) are analysed at the central control unit which optimizes the environmental conditions (e.g. temperature, relative humidity, etc.) to improve zonal thermal comfort dynamically.
At the core of aforementioned processes is a unique algorithm based on pattern recognition and machine learning, developed and tested together with domain experts at the School of Computer Science and Engineering. Test-bedding of this integrated approach is currently ongoing at ERI@N's office space with approximately 30 occupants and is aimed to improve thermal comfort as well as overall energy efficiency.