AI for HVAC System Optimization
Artificial intelligence to reduce energy consumption and increase the resilience of critical infrastructure
The Challenge
Partial and time-varying loads
Static programming logics
Variable weather conditions
Inefficient machine staging
Reduction in efficiency caused by operational wear
- Intelligent Optimization of Existing Systems Using AI Algorithms
- Integration of AI-Controlled Thermal Batteries for Dynamic Energy Management
Both solutions are designed for use in data centers, technology campuses, mission-critical infrastructure, and large industrial facilities, where energy efficiency and operational continuity are essential requirements.
Benefits for Data Centers and Large Facilities
Case Study
From Excess Energy to Operating Value
The logistics facility of a leading omnichannel retailer has a photovoltaic system that, during peak production periods, generates more energy than is needed. To make the most of this energy, a project was launched that involved installing a thermal battery integratedwith the HVAC system and implementing a smart control platform.
The goal is to increase self-consumptionof renewable energy, reduce grid draw, and improve the efficiency of existing chillers through predictive strategies based on energy data, weather, and thermal demand.
Customer requirement
To utilize this excess production and reduce grid consumption of the HVAC system, improving the efficiency of existing chillers without increasing installed capacity.
Project start
Starting from the existing thermal plant, we will implement a thermal storage system composed of 10 insulated tanks, booster pumps connected to the HVAC circuit, the refurbishment of the AHUs, and the upgrade of the control system managing energy shifting. The civil works have already been completed upstream; our scope includes piping, pumps, valves, AHUs, and the on-site integration of the control platform.
Activities carried out
Site coordination and team management; installation of field instruments for monitoring and control; mapping of variables and integration with the Amazon BMS; configuration of predictive and fallback logics for storage charge/discharge management, with real-time monitoring of PV data, weather conditions, and thermal demand.
Benefits achieved
Energy shifting of thermal loads to the most efficient time slots, with increased photovoltaic self-consumption and reduced grid draw. More stable HVAC operation (reduced peaks and on/off cycles), improved chiller performance thanks to the increased thermal inertia enabling charge and discharge operating strategies. The system provides historical traceability, diagnostics, and simplified maintenance, serving as a benchmark for efficiency and sustainability in the industrial sector.

