CorridorWatch

Using satellite imagery and Finnish open data to find changes near power lines that need a closer look.

  • 0 Raised
  • 63 Views
  • 0 Judges

Categories

  • This hackathon has categories available. Please select one if necessary.

Description

Idea Changes next to power lines are often noticed too late: storm-damaged forest, felling, digging or new construction. Satellites can already detect such changes, but operators get alerts, not answers. Most activity is legal and already recorded somewhere, yet nobody connects the change to the record that explains it. CorridorWatch adds that missing context layer: Detect. Explain. Prioritise. Alert.

Challenge Local sub-challenge "Resilient Finland – Space-enabled Critical Infrastructure and Command in a Hybrid Crisis" (Challenge 1): affordable, scalable detection of human-induced changes along transmission-line corridors.

How it works

  1. Map: cross-check power-line locations from OpenStreetMap and the National Land Survey and define the corridor.
  2. Detect: find forest loss and ground change with Copernicus Sentinel-1 radar and Sentinel-2 images.
  3. Explain: match each change with public records such as forest-use declarations and transport projects.
  4. Prioritise: rank the cases by whether they are explained, distance to the line, type of change and confidence. Uncertain cases go to a person.
  5. Alert: give the analyst a ranked queue with before/after images and a draft inspection ticket in plain Finnish. The analyst decides; a record is context, not proof of safety.

EU space technologies Copernicus Sentinel-1 and Sentinel-2 via the Copernicus Data Space Ecosystem, with ESA APEx processing tools.

Status Idea, architecture and candidate data sources identified; a first feasibility check matched a satellite-detected forest change with a forest-use declaration. The prototype is built during the hackathon weekend.

Hackathon goal One power-line corridor, chosen by data availability: detected changes matched with public records, a ranked queue in a map app, and a sample inspection ticket. We will validate the problem and the output with the Fingrid experts on site.

Team A team of four covering machine learning and satellite data, software and GIS, map and data sources, and business and pitch.