AI-Based Detection and Classification of Industrial Fires and Persistent Thermal Sources Using NASA FIRMS, OSM & Satellite Data
National Technical Research Organisation (NTRO) · Miscellaneous · Software
A tractable, well-scoped geospatial problem with all data free and linked — lean on the temporal-persistence signal to separate flares from fires, and be honest that FIRMS resolution limits facility-level attribution.
Data: firms.modaps.eosdis.nasa.gov/map
What it actually is
Satellites detect thermal hotspots from space, but systems like NASA FIRMS just report a hotspot without saying what it is — an industrial fire, a gas flare, crop burning, a wildfire or mining. The ask is a system that classifies these thermal anomalies by integrating the fire data with land-cover, industrial-facility databases and satellite imagery, so an industrial fire can be told apart from a forest fire.
What to build
A geospatial classification system that ingests NASA FIRMS thermal anomaly detections and, for each, pulls contextual layers — OpenStreetMap and industrial-facility locations, land-cover classification, and satellite imagery of the location — then classifies the anomaly into categories such as industrial fire, gas flare, thermal power plant, agricultural burning, mining or wildfire, distinguishing persistent thermal sources like flares from transient accidental fires by their temporal behaviour, and presenting the classified anomalies as an overlay on a GIS map with the ability to monitor persistent sources over time.
Smallest thing that wins the room
Load a region's FIRMS detections and show the system classifying them — a persistent hotspot at a known refinery location labelled as a gas flare, a cluster over farmland in season labelled agricultural burning, and a hotspot in forest labelled wildfire — each justified by its land-cover and facility context on the map.
How crowded this one gets
A guess, projected from the 2025 statements — the last year where both the submission counts and the winners were published.
Quieter than 37% of the 226 · #143 of 226 by expected field
A normal-sized field. Your idea has to be good, not miraculous.
Why: defence, intelligence and space bodies drew small fields.
This is a guess, not a fact
Nobody has published 2026’s numbers yet. This is an analysed estimate from last year’s pattern, so please do not take it as the truth — check the live counter on the SIH portal before you decide anything. The range covers the middle half of likely outcomes, so one statement in two lands outside it. Entry closes at 500 ideas per statement, so no range goes past that — a statement that reaches the cap fills and shuts rather than drawing an unlimited crowd. The model reads only three things a team can see before choosing — software or hardware, the theme, and what kind of body posted it — and those explain about a quarter of the variation in last year’s field sizes (R² 0.25 on held-out statements). Trust the band more than the number, and the ordering more than either. It cannot see how good your idea is, which is the part that actually decides it.
The scores
The number is the shorthand. The line under it is the reason.
Acceptance potential
4/5Underrated — all data is free and linked, the build is genuinely tractable, the classification is a real gap in FIRMS-based monitoring, and the geospatial-plus-thermal domain thins the field while the NTRO framing gives it clear purpose.
Feasibility
4/5NASA FIRMS is free and linked, OpenStreetMap industrial facility data and land-cover products are open, and the classification is a tractable problem of combining location context with the thermal detection's temporal persistence — all achievable with standard geospatial tooling.
Innovation scope
3/5The classification approach is fairly determined by the available context layers — location plus land cover plus persistence largely decides the class — so your room is in handling ambiguous cases and in the persistence analysis rather than in the concept.
Clarity
4/5The description names the data sources, the categories to distinguish, and the specific deliverable of separating industrial from natural fires with a GIS overlay, so the requirement is well defined.
Effort
HeavyFIRMS ingestion, contextual layer integration, the classification logic and temporal persistence analysis plus a GIS front end are focused, well-bounded pieces on open data.
Demo-ability
EasyClassified hotspots on a map, each justified by its facility or land-cover context, is immediately legible and the refinery-flare-versus-crop-burning contrast tells the story instantly.
In its favour
- Green flag: NASA FIRMS is free and linked directly, and the supporting OSM and land-cover layers are all open, so the entire pipeline runs on free data
- Green flag: Distinguishing a persistent gas flare from a transient accidental fire by temporal behaviour is a genuinely clever, buildable feature
- Green flag: Classified hotspots justified by facility and land-cover context make an immediately convincing map demo
- Green flag: It fills a real gap since FIRMS reports hotspots without classifying them, and the impact is concrete for infrastructure and safety monitoring
Against it
- Red flag: FIRMS resolution is coarse at 375 metres to 1 kilometre, so precisely attributing a hotspot to a specific facility is uncertain in dense industrial areas
- Red flag: Ground-truth labels for anomaly classes are scarce, so validating your classifier honestly is difficult and you may lean on land-cover as a proxy for truth
- Red flag: Agricultural burning is seasonal and location-overlapping with other classes, so seasonal context is needed or it is easily confused
- Red flag: Persistent industrial thermal sources and accidental fires can co-locate, so distinguishing a flare-up at a refinery from its normal flare is genuinely hard
What you will be writing
- NASA FIRMS thermal anomaly ingestion
- OpenStreetMap industrial facility data (Overpass)
- Land-cover classification (ESA WorldCover)
- Temporal persistence analysis for flare vs fire
- Sentinel-2 imagery context
- PostGIS + Leaflet GIS overlay
- Thermal remote sensing
- Industrial monitoring
- Geospatial classification
Prior art to read before you start
thermal anomaly classification · industrial fire detection from satellite · persistent thermal source monitoring
Analysed by Claude Opus. Every score above is a judgment call with its reasoning attached — kindly cross-check this against the official statement on the SIH portal before your team commits to it.