AI Rejuvenates Old Buildings for a Decarbonised Future

Ruchi Choudhary, Professor of Architectural Engineering, University of Cambridge

The Generation Gap in the Built Environment

Much like the generational divide in human societies, the built environment is split between old, inefficient buildings and the fast-changing AI-driven solutions that promise to decarbonize them. In the UK, nearly 80% of buildings were constructed before 1980. These buildings—many still reliant on gas boilers, outdated insulation, and inefficient heating systems—are now being asked to adapt to a new world of net-zero targets and renewable energy.

But change is never easy, especially for those set in their ways. For AI solutions to succeed, it must act as more than just a technical fix—it must embody the qualities that make intergenerational relationships work: communication, empathy, trust, responsibility, and perhaps even some playful interactions!

Communication: Speak to Aging Buildings

AI solutions face fundamental language barriers with older buildings. Think GenX lingo versus the Boomers! Where most AI solutions are being built atop the expectation of easy access to IoT sensors, automation, and cloud connectivity, pre-1980 structures operate on analogue systems, manual controls, and deeply entrenched operational habits.

To bridge this gap, AI solutions needs to act as translators between the old and new. We need non-invasive and accessible techniques to map energy performance and identify inefficiencies that are specific for a given building. We need techniques to ‘listen’ to older Building Management Systems (BMS) — decoding their thermal behaviour, energy demand cycles, and maintenance needs — before proposing radical changes. At the same time, we urgently need a renewal of facilities management as a profession that aligns with current technical advances in AI and machine learning.

Frugality please!

We are at the cusp of change. Generative AI is transforming our technologies, our decision-models, and our energy systems. This transformation is backed by a vastly improved landscape of data. Yet, data also comes with a carbon cost. Thus, it is not just in the interest of aging buildings that we need AI solutions that are data frugal, but also for the wider benefit of decarbonisation.

Empathy: Value Legacy

Resistance to change is common, not because progress is rejected, but because of disruption, loss of control, or unfamiliarity. The same applies to aging buildings. Replacing gas boilers with heat pumps, integrating smart meters, changing windows, or shifting to AI-driven demand response systems might make technical sense, but unless the process respects the existing structure, these changes will face pushback from planners, owners, and occupants. Older buildings hold architectural and societal value. AI technologies need to be aligned to aesthetic values and at the same time challenge outdated and prescriptive planning constraints.

We have some incredibly successful examples of architectural retrofits where the old and the new co-exist harmoniously, eg. the British Museum, St. Pancras International to name just two. Surely, we can harness the same level of creativity in bringing energy efficient solutions to the aging building stock.

Trust: Building Confidence in AI’s Decisions

Intergenerational relationships thrive on trust, which is earned over time. AI, too, must gain the trust of building managers and occupants by demonstrating transparency, reliability, and security. AI based solutions must present its recommendations in clear, actionable terms. For example, instead of saying, “Optimising Heating Ventilation and Air Conditioning (HVAC) system loads based on dynamic energy pricing,” AI could communicate, “Reducing heating at night can lower energy bills without affecting comfort.” Furthermore, trust grows when results are tangible. AI-powered energy optimisations should be implemented in measurable phases, allowing occupants to see real-world benefits, such as lower energy bills and improved comfort.

Playful discussions: Making Sustainability Engaging

One of the most underrated aspects of learning is playfulness, or gamified experiences. AI, too, can introduce an element of play in the process of decarbonising buildings, through adaptive and responsive interactions. Giving AI-driven building systems a relatable interface can make interactions more engaging, fostering a sense of connection between the building and their operators.