Personal comfort levels can differ across multi-occupant spaces, such as open plan offices, as they are influenced by several factors including activity, presence of co-occupants, and local effects of solar radiation.
Real-time data, or well-founded predictions, of the needs of occupants with regard to indoor thermal environment would enable more effective means for adjusting indoor conditions and reducing the incidents of productivity loss at the workplace. This requires a methodology for processing data from distributed, pervasive sensors in a way that balances the three-way conflict between the actual thermal conditions in the building space, the conditions that satisfy energy efficiency requirements, and the conditions that minimise productivity losses due to thermal discomfort.
This project is funded under Arup’s Global Research Challenge 2016-17.
Traditional approaches to building design and operation rarely address the aspects of indoor thermal comfort that are related to the personal preferences of occupants, variability of occupancy, and the outdoor microclimate around the building, and how these factors influence the productivity of a building’s occupants.
This research aims to develop sensor-fed learning algorithms to analyse the variability of the parameters influencing indoor thermal comfort, and to identify correlations between these (highly variable) parameters and the productivity of occupants.
Building upon available algorithms for associating sensor data with predictions of stochastic parameters in buildings, we will further develop prediction algorithms, particularly addressing the influence of outdoor microclimate, number of occupants and the duration the occupants are present. Adaptive learning capabilities will be incorporated with the aim of making the algorithms applicable to a wide variety of buildings and climates.
Low-cost, network-based sensors will be deployed in case study buildings to provide data for the development and validation of the adaptive occupant and microclimate prediction algorithms. Correlations will be developed between the indoor/outdoor microclimate parameters, the occupant preferences and their perception of productivity loss.
We plan that the findings from this work will be used to develop recommendations for a control strategy for new Arup offices at 151 Clarence Street in Sydney, Australia.
The findings from this project can be used to strengthen post-occupancy evaluations, commissioning processes and monitoring practices of buildings during their operation.
The proposed methodology and its adaptive control algorithms can be useful in satisfying the diverse and often conflicting needs of occupants and facility managers.
The research can inform ways of managing space in buildings, particularly the management of office spaces and the suitability of standard environmental control systems to support new ways of working, such as encouraging more spaces for collaboration, discussion and teamwork, and flexibility for allocation of desks.