About 50 percent. Riksbyggen was able to save a lot of heat energy in a month-long pilot project using artificial intelligence to improve heating. This is a relatively new and energy-efficient feature. The project is now being implemented in order to reduce energy consumption, not least the environmental impact these properties represent.
It started with a problem, but also a solution.
The problem is called operational optimization, and it relates to trying to keep the internal temperature in the building at an equal temperature, all year round, around the clock.
– The modern heating system relies on the presence of sensors that measure the outside temperature, as well as the presence of regulators that can increase and decrease the temperature of the elements, explains Jonas Holmberg, energy engineer at Riksbyggen.
The idea is that if the temperature outside falls below a certain level, the indoor temperature should be raised automatically, so that the indoor temperature is always comfortable. The only problem is that the system has a hard time planning ahead and taking into account that heating and cooling is a time-consuming process.
Say it’s a sunny summer day and the house is warm but at night the temperature drops below ten degrees for a while. Then the system wants to increase the heat even though those who live there just want to get a little cooler.
The problem could be solved if the engineer went several times a day and optimized the system for the external conditions applicable at the time, but it would be very expensive and time consuming. The solution emerged when Jonas Holmberg thought about how to automate this instead.
– We participated as a partner in a hackathon in October 2020 where programmers will improve a fictional city and its buildings, then I was surprised that it should not be difficult to do it in real life with the help of artificial intelligence.
The reason is that modern buildings have sensors that generate data around the clock. This information, along with a ten-day forecast from SMHI, allows the AI to calculate how hot or cold a building is within ten hours. Based on this, long-term thermoregulation is possible, something that can create a more balanced temperature as it avoids peaks and valleys, and at the same time reduces energy consumption.
– We started the pilot program at the end of March 2020, and during the months of April and May we saved a total of 15,500 kWh, which saves fifty percent of the heat usage. And that’s in a newly produced house that has been running silver in eco-construction and has been power optimized manually by us beforehand.
In total, Jonas Holmberg estimates that the technology could save about 20 percent of heat use over the course of a year, just on the heating side.
Real estate accounts for 40 percent of Sweden’s energy consumption, which means significant savings for the environment. In addition, it will be more fun to work with new technology as a power engineer, and since Riksbyggen already has a proactive IT department and management that listens and is not afraid to test new technology, this could have significant implications.
will continue to develop
The plan now is to develop more functions and train AI on several different types of characteristics, but also to continue developing more services where AI can help improve operations.
The first part of digitization was data collection, now it’s time to use the information to work and make changes, something that provides a better environment and a more pleasant workplace.
“Falls down a lot. Internet fanatic. Proud analyst. Creator. Wannabe music lover. Introvert. Tv aficionado.”