Overcoming Building Data Challenges with AI: Watch the Webinar
Now streaming: Part 1 of the “Demystifying AI for Buildings” webinar series.
In this 30-minute session, experts at Noda explore one of the biggest barriers to using AI in buildings today: messy and unstructured building data. It’s not the most glamorous problem, but it’s one that will quietly undermine every effort to optimize energy performance, improve building operations, and meet sustainability goals down the line.
Whether you’re new to AI in buildings or have experimented with a few tools already, this webinar offers a practical, grounded perspective on what really needs to be in place before any of the downstream benefits of AI can be realized – and how you can apply AI to solve your data problems, creating a scalable foundation for your energy and sustainability programs.
Why Building Data Is So Challenging
Commercial buildings are inherently complex. They’ve typically gone through multiple owners, vendors, and retrofits and upgrades. Each change can leave behind a trail of disconnected systems, inconsistent naming conventions, and partial documentation.
That creates challenges like:
- Multiple labeling schemes for the same equipment and sensor types across different buildings (or even within one site).
- Difficulty integrating systems without extensive custom mapping.
- Long deployment timelines for energy platforms.
The result? A massive backlog of unstructured and inconsistent data that makes it hard for teams on the ground to get reliable insights or take meaningful action. Enter: AI.
Before AI Can Optimize, Your Data Needs Structure
Our thesis is this: Before AI can optimize your building, it needs to understand what it's looking at. But AI isn't just for optimization – it can also significantly accelerate and scale the process of data normalization, mapping, and organization.
And it starts with something called an ontology.
Foundational AI Concepts, Explained Simply
We spent part of the webinar clarifying what we mean by “AI” in the context of building operations. Here’s a quick breakdown:
What Is an Ontology in Building Data?
An ontology is a standardized way of organizing and labeling building data. It defines how various components (like chillers, fans, thermostats, zones, and points) are named and related to one another.
Think of it like a common language and structure that allows both humans and machines to understand how things fit together.
If you’ve ever encountered a point labeled DATSP_2 and had to guess whether it referred to a discharge air temperature setpoint, an actual measurement, or something else entirely, you’ve experienced the lack of a consistent ontology firsthand.
At Noda, we use the Ontology Alignment Project (OAP) to apply a standardized, machine-readable model to incoming building data, so that different naming schemes and configurations can be understood in a consistent way.
What Is Generative AI in This Context?
When people talk about generative AI, they’re usually referring to models like ChatGPT that generate new text, images, or other content. But in the context of building data, generative AI is used to:
- Translate inconsistent point labels into standardized terms
- Predict likely relationships between equipment and systems
- Fill in missing metadata by inferring structure from context
These tasks are powered by techniques like Large Language Models (LLMs) and vector-based search, which allow the system to “understand” different ways of expressing similar information.
We also use a technique called Retrieval-Augmented Generation (RAG), which combines generative AI with reliable, domain-specific reference data – so we’re not just making guesses, but grounding them in what’s actually true about building systems.
What’s Changed: From Manual Mapping to AI-Aided Structuring
Historically, cleaning up building data was a slow, manual process that involved spreadsheets, engineering hours, and trial-and-error. In the webinar, we walk through what that process used to look like – and how it’s changed with new AI tools.
One example came from a data center, where over three different voltage/amperage labeling schemes were used within the same building. This made it incredibly difficult to determine what systems were connected to what equipment (and who was responsible for managing them).
By using AI-powered data modeling and the OAP, we were able to rapidly standardize and map those relationships across systems, providing a clean and structured foundation for further analytics and optimization.
Why This Matters Now
While cleaning and structuring data might seem like a backend task, it has huge implications for what you can do next.
- Shorter deployment timelines: Using AI to structure and organize data dramatically reduces the time it takes to get a new system up and running.
- Better analysis, more savings: You can’t run optimization programs or fault detection on messy data.
- Secondary datasets: Adding utility bills, emissions data, and real-time performance metrics only works if the foundational data is trustworthy.
- Scalability: A clean, structured data foundation will set you up for success as you seek to onboard new buildings or as equipment is added or changed.
If you’re planning to launch a carbon reduction initiative, integrate with ESG platforms, or roll out energy management across a portfolio, having a clean, normalized data model is mission critical.
How We Approach AI at Noda
Rather than starting with “AI” as the goal, we focus on what AI is being used for, and what needs to be in place to make that use case viable.
Here’s the basic sequence we walk through in the session:
- Ingest data from BMS systems, sensors, smart meters, and utility feeds.
- Normalize and enrich that data using the Ontology Alignment Project.
- Model relationships between equipment, points, and systems.
- Make the data usable for our downstream tools: analytics and FDD (Clarify), optimization planning, identification, and enablement (Conduct), and automation (Command).
The focus is always on solving the practical issue of “How do we get good, usable data out of what we’ve got?”
Takeaways
If you’re considering AI as part of your building operations or sustainability strategy, the message from this webinar is clear: AI doesn’t work without structure.
You don’t need to rip and replace your stack. You don’t need to be an AI expert. But you do need a plan for organizing and structuring your data – and tools that can make that process faster and more scalable.
This is the first in a 3-part series on AI for buildings. Part 1 focuses on getting your data ready. Future sessions will cover:
- AI for Optimization: How AI helps identify and prioritize operational energy projects
- Agentic AI: A look at the future of autonomous building operations
Want to Learn More?
If you’d like to talk through your specific data challenges or get a sense of how this applies to your building portfolio, we’d love to connect. Get in touch with our team here.
About Noda
Noda is a data and analytics company on a mission to make every building smarter, more efficient, and more sustainable. Recently ranked in the top 10 tech companies leading the charge on climate action, its AI-powered suite of products surface unique insights that empower real estate teams to reduce costs, decrease time spent on routine work, and find and act on opportunities to save energy and carbon. Discover how Noda's solutions can unlock the potential of your assets and accelerate the transition to net zero. Visit us at noda.ai to learn more.