Postdoc in machine learning for model predictive control with user interaction in buildings

Within the context of a Swiss Federal Office of Energy project named LUCIDELES (Leveraging User-Centric Intelligent Daylight and Electric Lighting for Energy Saving) the Human-IST institute from the University of Fribourg and the Energy Informatics group from the Idiap Research Institute join their forces in Building Interaction and Machine Learning for saving energy while preserving comfort in the built environment.

Starting from the observation that the increase of automation in buildings in order to diminish electric lighting and increase daylight exposure suffers from negative user perception, the project aims at studying the potential of a purposeful interplay between a user-centric interface design and an intelligent daylight and electric lighting control, in order to optimize comfort, promote its user acceptance and, in definitive, save electric energy.

By using machine learning for a fast and efficient model-predictive control of daylight and electric lighting, the candidate will participate to the setup of experiments. First in a controlled environment within two experimental office modules at the Smart Living Lab, and second in real office settings both at the Smart Living Lab and at the Idiap Research Institute for longitudinal user studies to evaluate acceptance, perception and satisfaction of users.

The ideal Postdoc candidate should hold a PhD in computer science with a specific focus on machine learning and a special interest for the application context of domotics. Alternatively, the candidate could hold a PhD in building physics with a specific focus in daylighting and/or electric lighting and a special expertise in applying ML techniques to this field.

The candidate should have expertise in some of the topics listed below :
• RADIANCE simulations for daylight and electric lighting
• non-linear surrogate models (machine learning)
• integration of real-time controllers for building components (e.g. electric lighting and blinds)
• web app development
• user studies planning and running
• monitoring of illuminance, luminance and glare evaluation

Additionally the candidate should have strong programming skills and be familiar with Python and the Linux environment.

The successful candidate will join (in equal parts) the Human-IST institute at UniFR and the Energy Informatics group at Idiap, under the supervision of both Dr Julien Nembrini and Dr Jérôme Kämpf. Appointment to this position is for an initial period of 1 year, with a possibility of a short extension.

Starting date is as soon as possible but open to discussion. The gross salary is in the Postdoc range of 80’000 CHF per year.

Interested candidates are invited to submit a cover letter, a detailed CV, and the names of three references through the Idiap online recruitment system:

Postdoc in machine learning for model predictive control with user interaction in buildings

All queries related to the advertised position can be sent to Dr Jérôme Kämpf or alternatively to Dr Julien Nembrini (julien.nembrini@unifr.ch).

The UniFR and Idiap place great emphasis on gender diversity and we particularly encourage women to apply.