About

Dead reckoning is how you navigate when the satellites go quiet: from your last known position, your heading, and your speed. The name is the thesis: this is a record of where, and when, aircraft were forced back onto their own instruments.

Why it exists

Several tools already show current GPS interference conditions: GPSJam, Flightradar24's jamming map, SkAI's spoofing tracker. They answer "what is happening right now." None of them keeps memory. There has been no scrubbable, multi-year archive that lets you ask: when did this region light up? how does today compare to its own normal? what was happening in the world that week?

DEAD RECKONING is that memory: per-hex baselines, anomaly scoring, named watch regions with analyst context, and a timeline of sourced events, all built from open data, with uncertainty shown as a first-class feature rather than hidden behind confident color.

How it works, briefly

Aircraft broadcast a navigation-integrity value (NIC) with every position. Where many aircraft simultaneously report degraded integrity, GNSS is likely being interfered with. We aggregate that signal per hexagonal cell per UTC day, compare each cell to its own history, and render the result. The full method, and its limits, is on the methodology page.

What it is not

It is not a jamming-attribution service. The map never asserts who caused an interference event. Region profiles summarize what public reporting attributes, with sources, and are labeled as analyst interpretation, not measurement.

Stack

A static site: no servers, no databases, no paid APIs. A Python pipeline (H3, streaming JSON) turns daily adsb.lol dumps into small flat aggregates; the frontend is vanilla JavaScript with MapLibre GL JS, h3-js, and D3, on an OpenFreeMap basemap. GitHub Actions rebuilds it nightly.

Roadmap

DEAD RECKONING is a living research program, not a finished dataset. Planned work: deeper historical backfill toward 2023; spoofing detection as a distinct layer, telling false positions from lost ones; a lead-lag study of whether interference spikes lead or follow documented events; a watchlist with a public calibration scorecard that grades its alerts against outcomes; and machine-readable per-region data feeds for other researchers.

Author

Built by Walker Robinson, using AI-assisted development under a verification and review process he designed. walker-robinson.com · GitHub

Data & licensing

Source flight data © adsb.lol feeders and partners, ODbL 1.0. Published aggregates: ODbL 1.0. Code: MIT. Basemap © OpenStreetMap contributors.