The Predictive Company's solution in Phase 2 of Ai4Cities is a predictive energy management system supported by AI, that learns the energy profile of a building infrastructure, to forecast its real energy demand with the end of an autonomous and optimized operation of the HVAC systems. It goes beyond the state of the art for two reasons:

Firstly, it is based on modelling the behaviour of the Coefficient of Performance of the equipment, focussing on optimization of the asset’s operation and its maintenance perspective. Secondly, the datasets of different buildings can be shared to impact over the efficiency of each single facility. This approach requires management of sophisticated AI tools and data management procedures that are integrated in the proposed solutions after years of research and experiments.

The solution works by combining predictive analysis (users can predict and optimise their future energy needs) with predictive control. This combination allows the software to focus on machine-operational optimisation and to detect even the smallest failure in the machines, before it becomes a serious issue.

The technology makes it possible to autonomously manage HVAC systems and detect early failures. This can potentially reduce energy consumption and CO2 emissions by 20%, and lead to happier employees due to a more optimised work environment. The Predictive Company believes that its solution has the potential to save 4 million tonnes of CO2 by 2024.