Adaptive predictive energy model

Using external (weather forecast, variation in energy tariffs) and internal (inputs and feedbacks from building’s users) information, the web platform lowers the building’s energy consumption and optimizes grid interactions. The platform defines the HEART toolkit operations, by applying a predictive-adaptive logic to guide the building system’s efficiency.

This type of logic makes best use of the passive features of the building. For example, based on predetermined or real-time boundary conditions, the logic can optimize solar gains for heating or the building envelope’s thermal mass to reduce both summer and winter loads.

The protocol uses a self-learning function, which improves timing by comparing simulation and forecast results with feedback from continuous building monitoring and from the users, in order to calibrate computational parameters.

Monitoring takes place through strategically positioned sensors (temperature, relative humidity, occupancy, CO2, etc.) which guarantee a continuous detection of the buildings’ efficiency. Using the model within HEART’s control logic means that the number of measurement points (along with the associated costs and installation works) can be minimized. These measurement points serve to verify and calibrate the mostly virtual, but highly reliable, monitoring. This information helps to manage operating conditions, energy consumption, mode of use, malfunction, maintenance needs, etc. The BEMS also enables direct interaction with occupants, who can keep track of the building’s performance and control it using dedicated apps on mobile devices (smartphones and tablets).

Thanks to these functions and technologies, HEART can transform an existing building into a highly efficient smart building. It optimizes the building’s energy performance and makes it extremely reliable, documented and transparent.


Forecast data used for the adaptive predictive energy model is provided by