At Itility, we strongly believe that data only creates value when it is tightly connected to operational processes. Entune brings that way of thinking into practice for energy management of buildings through BuildingAI: an approach where data, models and AI are used to continuously optimise how buildings perform. What started as intelligent energy control for buildings is increasingly developing into an ecosystem where technology, installation, management and day-to-day use come together.
Early work focused on optimisation algorithms that calculated better control-strategies, based on building data. Over time, richer data models and integrations with building systems were added, allowing those insights to translate into real control of physical assets.
That evolution is not driven by one breakthrough or algorithm.
It is the result of many deliberate choices made by multidisciplinary teams who take the real-world context of buildings seriously. Rather than more optimisation algorithms, the focus has shifted to understanding how buildings behave: physically, technically and in daily operation.
Technology serves a purpose. In buildings, that purpose is comfort, continuity and reliability. In public or social environments, that translates into a stable and pleasant indoor climate. In industrial or logistics settings, predictability and operational efficiency are often more critical.
Within Entune, data scientists and engineers work closely together to develop models that support these different realities. Building data is not treated as an abstract dataset, but as a reflection of physical systems such as heating, cooling and ventilation.
For example, large floor heating systems in warehouses behave very differently from traditional heating installations. Because of the thermal mass of the floor, heating up or cooling down takes hours. Optimising such a system is therefore not only about reacting to temperature changes, but about predicting how the building will behave over time and adjusting control strategies accordingly.
Modelling those systems requires more than data science skills alone; it demands an understanding of how buildings behave in practice.
Data-science-wise, the team works with so-called grey-box models: combining physics-based principles with data-driven optimisation. This makes outcomes explainable and more reliable, and helps prevent optimisation from drifting away from how a building actually works.
For many Itilians involved in Entune, the built environment was initially unfamiliar territory. HVAC systems, building management systems and installation logic are not typically part of a classic data science or software engineering background. That reality requires curiosity, asking many questions and, above all, avoiding assumptions.
This mindset reflects how Itility teams approach complex problems more broadly. Rather than assuming how systems should work, they take the time to understand how they are actually configured. That is especially important in buildings, where even owners or operators do not always have full insight into the technical setup. For reliable optimisation, that understanding is essential.
As a result, models are not only designed to be smart, but also carefully tested. Edge cases are considered upfront, exceptions are explicitly addressed and safeguards are built in to avoid unwanted outcomes. Optimisation should never be surprising to users – it should simply work better, every time.
Intelligent energy control only works when the entire chain moves along. That means close integration with physical assets such as installations and charging infrastructure, as well as alignment with the parties who design, install and maintain these systems.
By deliberately connecting software and hardware, genuine energy optimization becomes possible. Buildings no longer respond only to current consumption, but also take grid load, peaks and available capacity into account. This shifts building control from reactive to proactive.
At the same time, the role of installers is evolving. Beyond implementation and maintenance, they increasingly act as coordinators of intelligent, predictive building control. That shift requires solutions that are not only powerful, but also transparent and manageable in day-to-day operations.