Spotlight Interview: Building Atlas

In our latest spotlight interview, we spoke to Nick Taylor, CEO at Building Atlas.Nick shares how Building Atlas is transforming portfolio-wide decarbonisation by using AI and physics-based building models to deliver fast, scalable retrofit intelligence."Through that work it became obvious that real estate was massively underserved - people typically know that buildings are a significant part of their emissions (and energy costs), but they don’t really know where to start. Surveying buildings is just too slow and too expensive, and buildings are often tackled one-by-one."

Q1. What inspired the creation of Building Atlas, and what problem are you aiming to solve?

Olga and I had previously worked on a B2B supply chain decarbonisation platform at Google. Through that work it became obvious that real estate was massively underserved - people typically know that buildings are a significant part of their emissions (and energy costs), but they don’t really know where to start. Surveying buildings is just too slow and too expensive, and buildings are often tackled one-by-one.  

Q2. How does your AI approach differ from traditional building-by-building energy assessments?

We use a mix of AI and machine learning to build a physics-based living model of each building. We prefer this term to “digital twin” given how overloaded it is, and how people often promise more than they deliver.

The platform then optimises plans across the whole portfolio based on business goals - aiming to get buildings to EPC B or net zero; quick win projects that can be done without lease breaks; removing gas or anything else.

Climate risk and resilience is also an important factor for our customers, so we use a variety of sources of climate data to consider risks including flood, temperature and precipitation. Customers can see climate risks and interventions not just on their own but also the impact on retrofit projects.

The key difference to traditional assessments is that we can model massive portfolios in minutes, for a fraction of the cost - and that we keep that plan always up-to-date. 

Q3. Balancing speed, scalability, and modelling accuracy across large portfolios is a major challenge in retrofit intelligence, what approach does Building Atlas take to ensure reliable insights at scale?

We have a multitude of data sources that feed that model for each building, allowing us to build a rich model of energy consumption, in essence running the equivalent of EnergyPlus for every retrofit possibility in every building in a portfolio near-instantly.

We use AI to further refine the model - detecting anomalies in the data, reading building documentation and more.

Users give us feedback that they’re impressed with the accuracy of our modelling. The platform also gives clear guidance on accuracy and confidence, too – and, of course, allows our customers to override any value, if we got an attribute wrong. Since the energy model is internally consistent, changing an attribute, for instance, telling us that LEDs have already been installed, will immediately update the state of the building and the rest of the recommendations.

Q4. How do your modelling tools help clients balance carbon reduction goals with financial performance and capital allocation?

For every measure, the platform calculates cost, savings, and carbon reduction, as well as feasibility and constraints (such as grid capacity or lease breaks).

Users can set goals - aligning to CRREM, net zero, their SBTi targets or other custom goals, and budget constraints to determine their optimised plans. 

Q5. Finally, what’s next for Building Atlas as you continue to scale and develop the platform?

We’re enriching our energy monitoring and anomaly reporting, adding more features to our climate risk and resilience capabilities, and growing internationally.