Enhancing the diagnostic capabilities of buildings with AI

This is a repost of  an Infosquare news.


Gathered within a collaborative project funded by the NPR, three IT companies and three institutions from HES-SO are collaborating with Richemont, Groupe E, and Groupe E Celsius to explore optimization potentials in buildings using data science and artificial intelligence.

Buildings account for one-third of global CO2 emissions, yet their potential for energy optimization remains largely underutilized. Faced with climate change and energy supply challenges, the pressure is increasing to reduce energy demand while maintaining human comfort. The collaborative project DiagnoBat brings together the expertise of three research partners in artificial intelligence (AI), IoT, and optimization in the context of buildings, three industrial partners developing ‘data-driven’ solutions for buildings, and three companies providing concrete use cases.

Jean Hennebert, professor and head of the iCoSys institute at HEIA-FR, leads the project. He emphasizes the strengths of such a combination of skills in applied research. “For our institute, it is important to bring innovations from the rapidly evolving field of machine learning out of the laboratories and bring elements to the field. Collaborative projects are ideal for this: they provide effective bridges between academic research and technological partners capable of taking our ideas and prototypes and turning them into products for market deployment. Thanks to the ecosystem we have created with our partners in the DiagnoBat project, we have all the necessary ingredients to develop new capabilities with IT companies for the benefit of potential users.”

The goal of the DiagnoBat project is to create a dedicated computer platform for building diagnostics, integrating new data capture solutions, visualization, and the latest advances in AI for signal analysis. A spec

ific objective is to make the training methods of predictive ‘deep learning’ models more generic to facilitate their deployment on different types of buildings and signals. DiagnoBat builds on research conducted in this field over the past five years, specifically on the BBDATA, Facility 4.0, and Assainissement 4.0 projects of the Smart Living Lab in Fribourg.

At the end of the project, the developed innovations will be integrated into the commercial offerings of industrial partners, who will be able to sustain the prototypes for use cases benefiting field partners.

Project partners E-NNO, GradeSens, Groupe E, Groupe E – Celsius, Richemont, yord, HEIA-FR – iCoSys, HEIA-FR –Energy, HEIG-VD – IICT