ETT, the French pioneer in climate engineering, is already banking on AI to decarbonize buildings

At the forefront of this new industrial revolution, ETT has chosen to deploy AI for its customers and across all of the company's departments. Data, Data AI, generative AI, machine learning, automation—the Breton manufacturer is accelerating its shift toward the 4.0 era and championing a reasoned and responsible approach to AI.

For Antoine Millot, general director of ETT: "Artificial intelligence is a tremendous asset for accelerating the transformation of our models. For our clients, it addresses multiple challenges in the field: from better cost control on a construction site to optimizing the energy performance of a building, from logistics warehouses to swimming pools and shopping centers. However, it is essential to keep in mind the rightful place of AI, which remains a lever in the bundle of solutions for decarbonizing. We talk a lot about decarbonization, but let's not forget that the range of possibilities thanks to AI is much broader: sobriety, rational use of thermal equipment, resilience to climate shocks."

A roadmap in several phases, three of which are already underway:

Phase 1. Better control of equipment using data

Since 2024, ETT has been integrating, through CPE (Energy Performance Contract), a solution for managing and controlling thermal equipment which combines data to better analyze consumption habits and more closely control the use of machines, for example based on weather forecasts.

For example, ETT analyzes data to detect anomalies in consumption or comfort relative to the building's profile and sends alerts to refine its management. The goal is to ensure operation is consistent with the building and climate conditions.

Phase 2. Self-regulation of machines with AI data

ETT wants to go even further and is developing a system that allows the machine to regulate itself and consume less energy. ETT's goal: to enable energy savings through real-time anticipation by analyzing the building's internal and external conditions. This solution also ensures a longer lifespan for the machines by avoiding overstressing the components.

For example, in a shopping center, ETT controls the regulation of air conditioning systems based on the weather and the analysis of the building's reaction the day before. When a heatwave is reported, the equipment will automatically anticipate the triggering of passive cooling (by bringing in outside air for example) in order to delay the triggering of the air conditioning as much as possible while preserving comfort. Conversely, if a temperate episode is reported, the equipment will not start "unnecessarily" and as it would with conventional regulation. This translates, day after day, into significant savings over the year.

Phase 3. Real-time monitoring of equipment using digital twins and AI

ETT develops "digital twins," a virtual representation of the building. Using real-world data, it's possible to reconstruct the building's behavior, making it an ally in decarbonizing. A clear view of the quality of its insulation, for example, makes it possible to compare different sites for the same client, helping to direct investments to those most in need.

For example, mass data analysis allows the building's thermal profile to be reconstructed: its insulation and its reaction to external conditions, but also to the use made by its occupant (open doors, storage areas, etc.). These profiles make it possible to adapt the power of the equipment to be installed to its precise needs, but also to detect abnormal consumption depending on the weather.

For Antoine Millot, general director of ETT: "The consumption of thermal devices can represent around 50% of a building's energy consumption and represent 90% of the total cost (machine purchase and consumption). However, 85% of sites equipped with HVAC systems have overconsumption due to poorly optimized settings or uncontrolled modifications. Comparative studies demonstrate that optimized regulation of equipment using AI already halves their energy consumption! And we're only at the beginning..."

At the same time, ETT is launching predictive maintenance to anticipate breakdowns.

ETT is laying the foundations for the future deployment of AI in construction: the anticipation of breakdowns, in other words predictive maintenance, through the installation of communicating sensors that can detect the very first signals that lead to a breakdown.

This technology is currently being developed at ETT and initial tests are planned for 2025.

On a daily basis, AI is present in all business management processes

At ETT, AI is gradually infusing all of the company's departments. Its adoption can be seen in generative AI, data platforms, and even RAG (Retrieval Augmented Generation) document management. The company's philosophy is to increase performance, help streamline processes to save teams time and help them develop their skills.

Responsible AI, a charter of 'good conduct' adopted by the company

For Hervé Stéphanus, Deputy Managing Director of ETT: "The line between regulating AI and adapting businesses to integrate this new tool is sometimes complex. ETT closely monitors regulations, particularly the AI ​​Act, the first part of which came into force on February 2 and plays a key role in overseeing the development of artificial intelligence. At ETT, we believe that the use of AI must first and foremost be in line with the company's values. It is our responsibility to design AI that respects our standards and ethical principles."

ETT adopts a responsible AI strategy that is structured around founding principles, in compliance with the legal rules of the AI ​​Act, for example:

• Integrate AI into governance, with a code of conduct and use of AI, as referenced in the company's CSR report,
• Exclude the collection and use of “emotional recognition” information,
• Mandatory indication whether the content is generated by AI or whether a chatbot works thanks to AI,
• Train internal teams in the ethical, reasoned and compliant use of AI.