Coupling physics, humans and machines in holistic, resilient and automated syste

Our second theme 3 successful research call applicant, Dr John Craske, will be researching coupling physics, humans and machines in holistic, resilient and automated systems for building management.

Dr John Craske discussed his small call project

Background: Buildings account for 40% of our energy demands and are pivotal in determining our health and well-being. However, our ability to design and control their internal environments is sub-optimal from a systems perspective. In spite of advances in flow modelling and control technology, building management systems (BMSs) are relatively crude and incorporate ad hoc rules for control. Systems responsible for the control of heating, ventilation, air conditioning and lighting are rarely integrated. Typically, BMSs are not coupled with predictive models for building physics and do not adequately account for the fluctuating and evolving requirements of users.

Objectives: We will develop a software tool that allows students, researchers and practitioners to couple constituent models of an entire building system in a simple, flexible and informative manner. Our tool will provide an application programming interface to combine models of a building's climate, occupancy and management system. The tool will be adaptable to accommodate improvements and additions to the constituent models in the future.

Potential industry impact: Our tool will have the potential to (1) improve the energy efficiency and environmental conditions of buildings by providing a holistic framework for design and analysis; (2) increase productivity by automating the steps between model development and the implementation and calibration of a BMS; (3) increase the awareness and competency of practitioners and researchers in the future by providing an educational tool.

Update: The project was undertaken during the summer of 2018 by two UROP (Undergraduate Research Opportunities Programme) students. One of the students focused on writing a software module to convert IndoorGML data, describing the layout of buildings, into a graph that can be used to model and simulate the environment of a building. The other student focused on developing and analysing the simulations.

Our work contributed to the development of an EPSRC proposal on data assimilation and has provided a starting point for further work relating to building control. We are currently working with real-time observational data from the College's building management system to test our modelling strategy and we are developing techniques to incorporate our approach within models of urban flow systems with other members of the Department. The software we developed will form the basis of an MEng research project on data assimilation for buildings and further MSc projects on the analysis of flow networks.