Highlights from the 133rd OGC Member Meeting (Boulder, Oct 28–30)

We had a great week in Boulder at the 133rd Open Geospatial Consortium (OGC) Member Meeting.

Our talks

Innovation Summit – Talking to the Planet: Natural Language × Digital Earth for Disasters

Co-presenters: Dr. Steve Liang & Dr. Erin Li

We framed a simple idea: what if we could ask the planet a question in natural language about floods, fires, earthquakes, and get a trustworthy, auditable answer? Our answer is to pair DGGS (a standardized, equal-area global grid with multi-resolution cells) with AI agents to make disaster data explainable and traceable.

We walked through the motivation (multi-source, multi-format, multi-CRS, multi-semantic data chaos), introduced DGGS as the common spatial/temporal/scale/semantic reference, and showed how a multi-agent pipeline (Intent Parser → Geo Resolver → Data Query) turns a plain-English question into DGGS-aware SQL and a sourced, cell-referenced result. Erin also demoed our methane query prototype to illustrate the pattern end-to-end.

The room was full and engaged, and we received good questions centred on the agentic architecture, implementation details, and future research plan.

DGGS Domain Working Group – Ask, Retrieve, Analyze: A Multi-Agent DGGS Framework for GenAI-Driven Methane Data

Presenter: Dr. Erin Li

This talk went deeper into the architecture and our two-part contribution, including data harmonization and multi-agent orchestration.

A notable theme across the meeting: DGGS is having a moment. DGGS-related talks appeared in multiple sessions, and the hallway conversations echoed the need for spatial/temporal/scale/semantic standardization to make AI results reliable.

EmissionML SWG

At the EmissionML SWG, Dr. Steve Liang walked through what’s in the current slide deck: the purpose of EmissionML as an abstract ontology and data model for emission events (extensible to specific types such as MethaneML), the roadmap and timeline toward OGC/ISO reviews, a progress update (use-case catalog, draft UML, and the Discussion Paper in GitHub), and touchpoints with core OGC specs (EDR, Moving Features, TimeSeriesML) plus plans for a consistent JSON encoding and SDKs. He also highlighted outreach and coordination items (e.g., upcoming training and partner engagements).

The discussion was highly engaging, especially on the ontology. Participants, including colleagues from other OGC working groups with deep ontology experience, offered concrete feedback on concept alignment, JSON patterns, and how EmissionML can interoperate cleanly with existing OGC standards.