The ongoing electrification of the European heating systems means that heat pumps are here to stay. It is projected that at least 10 mio. new units will be installed towards 2030 – a more than doubling of the existing installation base, imposing a significant strain on the energy-grid.

Both air-to-water and geothermal heat pumps are some of the increasingly popular substitutes for expensive oil and gas furnaces in private homes, however, many heat pump systems are limited in their ability to automatically adjust temperatures in response to weather forecasts, energy prices or the energy needs of the inhabitants.

By combining a digital twin, simulation, and AI technology, CEDARs system can provide significant cost and energy savings – without the loss of comfort. To the benefit of both the environment the budget.

COMBINING AI AND HEAT PUMPS FOR PREDICTIVE HEATING

With their innovative technology, CEDAR gives a new, modern AI-accelerated control to both new and existing heat-pump installations.

Contrary to regular heat pump control, CEDAR utilizes the future for controlling. By using their digital twin technology, they can predict the effect of changes in weather, electricity prices, and residential behavior – and change the heat pump operation accordingly. As the heating system becomes cost-aware, it allows for a reduced heating cost by shifting the energy use periods leading to better utilization of green energy.

The digital twin furthermore improves the comfort of the home. Via personalized residential digital twins CEDAR can ensure that there is hot water available when needed and that rooms with large windows are not overheated during sunny days.

The aim of CEDAR is to do cost-based fine-tuning automatically using predictive digital twins trained and personalized for each individual house. Preliminary results of the technology demonstrate cost savings of up to 50%, equating to hundreds of euros for a single-family house.

A GROWING MARKET – WITH A NEED FOR INTELLIGENCE

By utilizing advances in the automatic generation of predictive digital twins using gray-box and black-box modeling, CEDAR aims to address an unmet need for intelligence and flexibility in an ever-growing market.

CEDAR’s state-of-the-art machine learning and AI optimization technology is packed in an affordable, subscription-based edge-computing platform, easily accessible via the cloud.

While the primary customer base is homeowners with existing or new heat-pump installations, the platform furthermore has the potential to be expanded into management of energy-systems with large thermal capacity, e.g., swimming-pools, office-buildings etc. and secondary, the energy-grid sector for providing grid-stability function. This relies on a sufficiently large pool of primary customers which enables shifting loads to reduce over-utilization of the existing infrastructure.

THE COMMERCIALIZATION PROCESS

CEDAR is founded in 2024 by CEO and Associate Professor Peter Gjøl Jensen from the Department of Computer Science at Aalborg University. The team furthermore includes CTO and co-founder Kenneth Yrke Jørgensen from Aalborg University.

To strengthen the research team towards the establishment of the spin-out, the AAU Technology Transfer Office has supported the team throughout the process including the development of a business model, two successful applications for an InnoExplorer and InnoFounder grant, and match with external investors.

In 2024, the technology was licensed to CEDAR to accelerate the journey towards market.