Buildings account for a significant share of Sweden’s total energy use. Ventilation, heating, and electrical systems are often controlled by fixed schedules that do not adapt to how a building is used, to changing weather conditions, or to interactions between systems. This results in avoidable energy waste.

CAIR develops AI-based methods to control ventilation, heating, and electricity use in a more flexible and energy-efficient way. We build on earlier research at Örebro University and ÖBO, where AI-based ventilation control in a single building showed clear energy savings. In CAIR, we take the next step: investigating how AI systems across multiple buildings can learn from each other and thereby improve control in new and varied settings.

We collect and structure operational data from real buildings and develop models that can predict needs and adapt control in real time. We also test how knowledge can be shared between buildings while preserving each building’s local adaptation. Solutions are designed to work with existing building management systems.

Results target property owners, technology providers, and policymakers who want to reduce energy use in existing buildings without reducing indoor comfort.