Next-gen HVAC could use quantum computing

Given that residential heating, ventilation and air conditioning (HVAC) systems make up a large proportion of energy usage in buildings, there’s an increasing need for building managers to invest in technology that monitors and optimises energy use.

Occupancy-sensing HVAC control is one promising option, capable of providing 20–50% energy savings in homes. However, this type of tracking technology suffers from long payback times and privacy issues — and it isn’t designed to provide optimal comfort for residents. There is still a need for advanced technologies that can not only control energy, but also help to regulate indoor air quality.

To meet these expectations, scientists have recently turned to intelligent control methods such as quantum reinforcement learning (QRL) based on quantum computing principles. It’s an approach that can notably accelerate the machine learning process as well as handle the complexity of real-world building dynamics.

A group of researchers from South Korea, led by Sangkeum Lee, Assistant Professor of Computer Engineering at Hanbat National University, have presented what they claim to be the first demonstration of continuous-variable, quantum-enhanced reinforcement learning for residential HVAC and home power management. Their findings have been published in the journal Energy and AI.

Highlighting the novelty of the work, Lee said, “Unlike conventional reinforcement learning techniques, QRL leverages quantum computing principles to efficiently handle high dimensional state and action spaces, enabling more precise HVAC control in multi-zone residential buildings.

“Our framework integrates real-time occupancy detection using deep learning with operational data, including power consumption patterns, air conditioner control data, and external temperature variations.”

The proposed technology integrates features such as multi-zone cooling — to control the temperature of individual zones in a building — and clustering, to group similar data points and adjust cooling. In this way, a single controller jointly optimises comfort, energy cost and carbon signals in real time.

The researchers performed simulations based on real-world data from 26 residential households over a three-month period. They found that QRL HVAC control outperformed common machine learning algorithms such as deterministic policy gradient and proximal policy optimisation, achieving reductions in power consumption and electricity costs while still maintaining thermal comfort.

QRL HVAC control comes with many more benefits, according to the researchers. It is retrofit-friendly and works with standard temperature, occupancy and CO₂ sensors, as well as common HVAC equipment and thermostats. It can also handle uncertainty, such as noisy forecasts on weather and occupancy, as well as device constraints. Additionally, it has a generalisable framework that can be extended from apartments to small buildings and microgrids.

“It can be utilised in smart thermostats and autonomous home energy management systems that co-optimise comfort, bills and emissions without manual tuning — and rooftop photovoltaics and home battery scheduling,” Lee said.

“Our framework is also useful for utility demand-response and time-of-use programs with automated control.”

QRL-based HVAC control could be applied at community or campus scale through grid-interactive efficient buildings and virtual power plants (VPPs). (VPPs consist of groups of distributed energy resources — such as household solar panels and batteries — that help stabilise the grid by coordinating the use of renewable energy.)

As hardware matures in the coming years, the quantum approach could facilitate faster training for complex multi-energy systems such as HVAC, electric vehicles and energy storage systems. In the long term, the researchers expect their work to pave a way towards standardised secure controllers that can be certified and deployed on a large scale.

The paper can be read at DOI: 10.1016/j.egyai.2025.100541.

Image credit: iStock.com/jamesteohart