A Simpler Path to Occupant-Centric Control in Existing Buildings

Improving building energy performance should not come at the expense of the comfort of the people living or working in it. Occupant-centric control (OCC) has emerged as a promising strategy to reduce building energy consumption while maintaining occupant comfort. It is a building energy and temperature management strategy that uses real-time data to match occupants’ needs. 

Although promising, it has not been implemented in many spaces because it often relies on advanced sensing and computational infrastructure that is not easy to implement and often impractical for existing buildings. 

These concepts and a solution are explored in a new article published in Energy and Buildings co-authored by Sicheng Zhan, Assistant Professor at the University of Hong Kong and former postdoctoral associate at MIT; Audun Botterud, Principal Research Scientist in the MIT Laboratory for Information and Decision Systems; Jeremy Gregory, Executive Director of the MIT Climate and Sustainability Consortium; and Leslie Norford, Professor in the Department of Architecture at MIT. They present an integrated rule-based OCC solution that does not rely on advanced sensing or computational infrastructure and is compatible with existing building management systems. 

Support from Joseph Higgins, Vice President for Campus Services and Stewardship, and Francis Selvaggio, Senior Engineer Building Management Systems in the Department of Facilities, enabled the research team to use the MIT campus as a living laboratory in which their approach could be tested.

Jeremy Gregory highlighted the unique nature of the project, “this project illustrates the benefits of collaboration between research and operational teams in a university. Many building technology researchers publish work that is theoretical. We were fortunate enough to collaborate with the MIT Department of Facilities to test our ideas on the MIT practice. Now we can implement them to reduce energy consumption on campus.” 

The framework was deployed in two buildings at MIT for 10 months and achieved more than 40% energy savings compared with the baseline static nighttime setback control. A key finding is that there is potential for energy reduction in ventilation systems during heating seasons, which can be particularly challenging. Furthermore, the work demonstrated that OCC does not have to be complex – it can be implemented into existing building management systems in a relatively straightforward manner, while still being effective. 

The research team also addresses the trade-offs of AI and machine learning-based approaches through simulation-based comparative experiments. The work showed that their proposed OCC framework had results comparable to model predictive control, but with a lower implementation complexity, thereby making their framework more accessible.

Lead author Sicheng Zhan sees lots of opportunities to advance the work in the future. “Larger-scale implementations at MIT will amplify the energy savings and shed light on how we can improve implementation of the approach. We are also exploring the use of cutting-edge AI models to support the diagnosis and optimization of building system performance with minimal expert interference. Together these steps could make OCC more accessible to the people who operate and occupy buildings.”

Read more here