Monday Live Recap: Ontologies, AI, and the Future of Building Data
This week’s MondayLive! the session explored whether ontologies still matter in the age of AI, how they shape data interoperability, and what the future holds for semantic modelling in buildings.
Key Takeaways from This Week’s Discussion
1. The Role of Ontologies in AI-Driven Buildings
- Ontologies provide structure—AI thrives on clean, contextualized data, and ontologies help define relationships between systems, points, and metadata.
- Human-AI communication—While AI can process unstructured data, ontologies allow humans to interpret and interact with AI-driven insights.
- Metadata matters—Beyond naming conventions, ontologies help enrich data with contextual details (e.g., location, function, dependencies).
2. AI vs. Ontologies: Complementary or Conflicting?
- AI doesn’t need ontologies—Advanced AI can derive patterns without predefined structures (e.g., Keith’s example of AI optimizing set points without knowing what a “thermostat” is).
- But humans do—Ontologies ensure stakeholders speak the same language for maintenance, design, and collaboration.
- Potential constraints—Forcing rigid ontologies on AI might limit its ability to discover novel optimizations.
3. The Semantic Tiger Team Kickoff
Last week, the Semantic Tiger Team launched, bringing together 39 industry leaders to align disparate ontology efforts (Haystack, Brick, DBO, etc). Hosted under the Linux Foundation, this cross-industry initiative brings together the architects of major ontology standards to:
The Semantic Tiger Team isn’t creating another standard—it’s building the Rosetta Stone for smart buildings. Success would mean:
- Map the Ontology Landscape Document overlaps/gaps between Brick, Haystack, ASHRAE 223P, and RealEstateCore
- True plug-and-play interoperability – Mix/match systems as easily as linking websites
- 70% lower analytics deployment costs – AI tools auto-discover and interpret data
- Future-proof investments – Buildings gain long-term flexibility to adopt new technologies
The team aims to deliver its first interoperability framework by Q4 2024, with early adopters including major REITs and global tech firms. As Steve Bushby noted:
“This isn’t about replacing standards—it’s about making them work together so owners can finally get value from their data.”
For building professionals, the message is clear: The age of semantic interoperability is coming—and it will reshape how we design, operate, and upgrade buildings.
4. Will Ontologies Survive in an AI-Dominated Future?
- Short answer: Yes, but they’ll evolve.
- For now, ontologies remain critical for human decision-making.
- In the long term, AI may develop its own “language” (like tokens or mathematical models) but will still need to translate findings back to humans.
- Metadata standardization is the next frontier—without it, AI lacks context for deeper automation (e.g., predictive maintenance).
5. The Big Question: Can AI Fix the Ontology Mess?
- Instead of manual alignment, could AI auto-normalize competing ontologies?
- “Feed all existing models into an AI and ask for the optimal unified version.” — Keith
- Challenge: Many ontologies aren’t yet open/public, slowing progress.
Final Thoughts
The debate isn’t whether ontologies will persist but how they’ll adapt. As AI reshapes building operations, the industry must balance structure (for humans) and flexibility (for AI). The Semantic Tiger Team’s work could be pivotal—if it focuses on practical, AI-aided solutions rather than reinventing the wheel.
What’s next? Watch for updates on the Tiger Team’s progress—and whether AI steps in to streamline the process.
About Monday Live: A weekly forum exploring the future of smart buildings, IoT, and AI. Hosted by Enzo Bddier and featuring insights from industry leaders. Views expressed are personal, not organizational.
Next Week: Tune in for deeper dives into AI-driven building controls
#SmartBuildings #AI #Ontologies #SemanticWeb #BuildingTech
