DiagnoBat – Diagnosis of Optimization Potential in Buildings

DiagnoBat Diagnosis of Optimization Potential in Buildings
Realization
  • HES-SO Fribourg
  • Frédéric Montet
  • Alessandro Pongelli
  • Vincent Magnin
  • Prof. Jean Hennebert
  • Prof. Jean-Philippe Bacher
Keywords
  • Building management
  • Building performance monitoring
  • Data science
  • Big data
  • Artificial intelligence
  • Machine learning
Our skills
Strong expertise in machine learning and building energy consumption
Partners

Institutes

  • iCoSys Institute
  • ENERGY Institute
  • HEIG-VD, IICT

Companies

  • E-NNO
  • GradeSens
  • Groupe E
  • Groupe E – Celsius
  • Richemont
  • yord
Funding
NPR – PromFR
Schedule
2022 – 2024

Context

Buildings account for one third of global CO2 emissions and have an underutilized potential for energy optimization. With climate change and emerging issues related to energy supply, there is increasing pressure to reduce the energy demand of buildings. Beyond this aspect, it is also about maintaining human comfort and thus finding the best compromises.

Recent developments in data science and IoT provide new opportunities in the field of data acquisition, processing, visualization, and analysis from buildings. More specifically, recent advances in artificial intelligence (AI) can now be applied to building data with improved analytical and predictive capabilities.

Objectives

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

A specific objective is to make the training methods for predictive deep learning models more generic in order to facilitate their deployment across different types of buildings and signals. DiagnoBat builds on research conducted in this field over the past 5 years, and specifically the projects BBDATA, Facility 4.0, and Assainissement 4.0 from the Smart Living Lab in Fribourg.

From Applied Research to Industrialization

The DiagnoBat project brings together the expertise of 3 research partners in AI, IoT, and optimization in building frameworks, 3 industrial partners developing data-driven solutions for buildings, and 3 field partners providing concrete use cases.

DiagnoBat will enable industrial partners in the project to adopt the developed solutions through concrete use cases in order to:

  • Represent and analyze building monitoring data from a “building intelligence” perspective.
  • Detect anomalies in the operation of technical installations for predictive maintenance.
  • Identify interventions on technical installations and renovations, prioritize renovations.

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