Traditional heat and ventilation air conditioning (HVAC) systems regulate temperature in rooms and buildings by relying on temperature sensors and consolidated rule of thumbs based on, e.g., room occupancy. Quite ironically, this paradigm does not consider at all the actual human factor in the room, which is exactly the reason why temperature management exists in the first place. Recently, alternative control approaches based on human feedback have been explored, for instance asking occupants to communicate a three-level satisfaction score (cold/fine/hot) via an app on mobile phones. This, together with the availability of personal comfort systems such as heated or cooled chairs and desk fans, can greatly improve building efficiency by simultaneously adapting to the occupants' needs and reducing energy consumption by acting on small personal devices (e.g., cooling a person's desk space by switching on a desk fan) rather than turning on room- or building-wise HVAC.

Although promising, temperature regulation with human feedback is an active, underexplored research field. In this thesis project, you will be tasked to test human-in-the-loop feedback controllers on a realistic building simulator (IDA-ICE interfaced with MATLAB [1]), to carefully assess the benefits and the best control strategy in a realistic setting. You will start from existing simulation code developed in a previous thesis for a preliminary study on this subject [2]. Satisfactory outcomes will be submitted for publication on a high-impact journal.

Internship at KTH is expected to integrate this thesis project into a broader research collaboration. Please be aware that scholarhips/reimbursement might be available but is not guaranteed; please ask me or prof. Luca Schenato (schenato@dei.unipd.it) for any opportunities to financially support the internship abroad.

[1] https://www.equa.se/en/ida-ice
[2] https://thesis.unipd.it/handle/20.500.12608/98912