Carleton University
IN-4: Innovation 4

The Canal Building, one of Carleton University's newest academic buildings, is being actively used as a living laboratory for which controls and operational strategies are being continually developed and tested.

The Human Building Interaction Lab at Carleton is engaged in three main research and development projects as follows. 1) Adaptive/learning controls: Controls are being developed that learn occupant habits and preferences for comfort and the thermal properties of offices using a sparse set of standard sensors and actuators. For example, in one application the new control algorithm learns when an individual occupant is likely to arrive and depart on a given day of the week. Meanwhile, the controls are also learning how long a specific room takes to heat up from nighttime setback to daytime setpoint based not only on envelope properties but also heating system capacity and room contents (e.g., thermally massive components like furniture and books). Combining knowledge of occupants habits and room characteristics, the heating and cooling setpoints can be optimized for energy savings and comfort. This project is funded by Delta Controls, Regulvar, and NSERC. 2) Data visualization, diagnostics, and BIM: The Canal Building's automation system yields vast amounts of data (e.g., room temperatures, light states, heating and cooling energy use, etc.) that largely go unexploited. This project involves presenting performance data in engaging and revealing ways to identify opportunities for improvement to building operations. Sample projects include using building information modelling (BIM) and building performance simulation to assess different building-scale control strategies, data mining operations data for fault diagnosis, and visualizing energy flows from the point where is enters campus to the point it leaves the building. This project is funded by Autodesk and NSERC. 3) Occupant behaviour observational studies: Occupant behaviours is normally treated very coarsely and simplistically in existing predictive building simulation in design practice. But there's a wealth of information that can be obtained by monitoring various aspects of occupant behaviour (e.g., light switching, office equipment energy use, blind movement) that can be incorporated into the design process of new buildings. Using observed events of many occupants from raw building controls data over months or years, statistical models are being developed and implemented into advanced building simulation tools for use by the simulation community. These models can help improve predictive accuracy and inform designers of building design strategies that are more comfortable. This project is funded by NSERC and Natural Resources Canada. In closing, the Canal Building has proven to be a valuable hands-on training opportunity for students and a powerful facility for testing novel control algorithms. Once the research is more developed, it will be extended to other Carleton buildings and even off-campus.