Quantum Computing could Revolutionise Low-Carbon Building Management
Buildings, for all their bricks and mortar, are voracious consumers of energy. According to the International Energy Agency (IEA), they account for nearly 40% of global energy use and more than a quarter of energy-related CO2 emissions. That’s not a statistic to be taken lightly. But as the world hurtles towards net-zero targets, a fresh wave of technological innovation is giving rise to smarter, greener buildings. At the heart of this change? Quantum computing.
In a ground-breaking study published in Engineering, researchers from Cornell University, Akshay Ajagekar and Fengqi You, present a bold new strategy: merging quantum computing with model predictive control (MPC). This fusion aims to optimise building operations, improve energy efficiency, and slash carbon emissions—all while maintaining occupant comfort.
“We developed an adaptive quantum approximate optimisation-based MPC strategy tailored for smart buildings with energy storage and renewable generation,” explained the authors. “Our approach addresses the growing need for efficient, flexible, and low-carbon energy management.”
Smart Buildings Meet Quantum Brains
The core of their approach lies in the Quantum Approximate Optimization Algorithm (QAOA), which they refined using a learning-based parameter transfer method. In layman’s terms? It’s a way of fine-tuning quantum circuits so they adapt in real time to changes in a building’s energy profile, external weather conditions, or grid demand.
They paired QAOA with Bayesian optimisation and Gaussian processes to predict the best starting parameters for quantum circuits. This sharply cuts down on trial-and-error computations, reducing time and computational resources.
By framing the MPC problem as a Quadratic Unconstrained Binary Optimisation (QUBO) task, they tapped into the true strength of quantum systems: solving ultra-complex problems with countless variables far quicker than traditional computers.
Real-World Testing on Campus
Ajagekar and You didn’t just theorise their way through the study—they put their method to the test. Using real-time data from two buildings on Cornell’s campus, they compared their quantum-enhanced MPC against standard deterministic MPC models and quantum annealing.
The results were eye-opening:
- A 6.8% increase in energy efficiency over deterministic MPC.
- A 41.2% annual reduction in carbon emissions, achieved by better orchestrating battery storage and on-site renewables.
Not only did the system excel in terms of performance, but it also proved nimble. “The proposed strategy could dynamically adjust heating and cooling loads based on outside temperatures, ensuring both energy savings and thermal comfort,” the researchers noted.
Learning and Adapting with Each Iteration
While most systems struggle with the initial learning curve, the Cornell model powered through it. The QAOA needed more iterations early on, but as the system learned and adapted, it rapidly improved, eventually outpacing quantum annealing in performance.
This is where the learning-based parameter transfer paid off. By allowing knowledge from one scenario to inform another, the system became more efficient with time. That adaptability is crucial in a world where building conditions and climate variables shift daily—if not hourly.
Addressing the Limitations
Of course, it wasn’t all smooth sailing. The research team admitted their model was built around relatively simple building energy systems. In more complex environments, the sheer number of variables could potentially swamp QAOA’s current capacity. Plus, while the algorithm adapted well, it didn’t directly account for uncertainty quantification.
That said, they see this as an opportunity rather than a setback. “Incorporating uncertainty metrics and testing the strategy in larger, more varied buildings will be crucial next steps,” they wrote.
Opportunities for Expansion
What makes this development particularly exciting is its scalability. Quantum-enhanced MPC could be tailored for:
- Commercial office buildings with intricate HVAC systems
- Industrial facilities using large-scale process equipment
- Hospitals and data centres requiring round-the-clock energy efficiency
And it doesn’t stop there. Integrating real-time carbon intensity metrics into the model could make it even more dynamic, choosing not just the most energy-efficient, but the most carbon-efficient options depending on the grid mix.
“Extending the framework to more complex control scenarios and refining quantum algorithms could open doors to widespread, real-world applications,” the authors concluded.
The Bigger Picture
This research fits into a growing movement. Global leaders are already investing in smart infrastructure that supports digital twins, AI-powered automation, and now, quantum optimisation. Governments and corporations alike are hunting for scalable strategies to meet climate targets without disrupting economic growth.
Cornell’s innovation offers a tantalising blueprint for what’s to come: a future where buildings don’t just use less energy—they think critically about how, when, and why they use it.
A Promising Path Forward
Quantum computing is no longer a far-off concept reserved for labs and whiteboards. It’s stepping into the real world, bringing a new level of intelligence and adaptability to the built environment.
Sure, there are hurdles to clear. Complexity and uncertainty remain significant challenges. But the success of this pilot project signals something powerful: that with the right mix of quantum smarts, predictive control, and renewable integration, the buildings of tomorrow can become key players in the fight against climate change.
“We’re just scratching the surface of what’s possible,” said Ajagekar and You. “Quantum optimisation offers a whole new frontier for sustainable energy management.”
