Two Swedish municipalities are already using AI for thermal control

The system learns and adapts over time

Växjö and Eslöv municipalities in Sweden are pioneers in terms of the introduction of artificial intelligence technology for the use of thermal regulation.

The AI technology has been designed by NODA Intelligent System with a view to coordinate the heating and cooling processes of a building more efficiently.

Unlike traditional control systems, this AI solution is self-learning and proactive. Based on available measurement data, the system teaches itself over time and gets better, which means that it can adapt to the property’s needs in the future.

The system is more sustainable than people’s behaviour can be

The Växjö municipality, for example, sees the value of digital solutions within the energy and climate area. It has entered an innovation partnership with a consortium in order to achieve a better indoor climate in three schools and at the same time save energy.

“The biggest challenge is not to technically achieve a good indoor climate. The difficulty is to get people to comply with what is the best behaviour to make the most of the solution. In Växjö it can, for example, be about the students who are on the school premises. And the smarter digital systems you build, the closer you tend to get to people and the better your system must be at creating a relationship between people and technology,” explained Christian Johansson, CEO of NODA.

NODA also has Eslöv Municipality as a client, where about 45 buildings are being connected to save energy, improve the indoor climate for the property owner and create thermal flexibility for the local energy provider Kraftringen, so that they can optimise production and distribution.

The system will optimize energy use, while at the same time giving tenants a more comfortable and more even indoor climate. A common problem that can now be reduced is the uneven room temperatures that arise during rapid weather changes, as ordinary heating systems regulate too slowly. 

To enable automatic learning, temperature control and troubleshooting, a wireless temperature sensor is placed in each apartment to correct the control if any apartment deviates from the target.

When added heat is used more efficiently, heat consumption is reduced. This is important not least because the district heating delivered to Eslöv is not really enough during the coldest days of the year. 

In those cases, there is a local bio-oil boiler that is started up. A possible goal of the project is that this boiler should be able to be closed without compromising the available district heating capacity for further expansion in Eslöv.