Key points
- Project overview: World Monitor (GitHub: koala73/worldmonitor, ~23.3k stars, AGPL-3.0) is a free, open-source global intelligence dashboard. Its homepage is https://worldmonitor.app.
- Four variants serve different use cases:
- World Monitor — geopolitics, conflicts, military posture
- Tech Monitor — AI/ML, startups, cybersecurity
- Finance Monitor — global markets, central bank activity
- Happy Monitor — positive news and human-interest stories to counter information anxiety
- Data coverage:
- 150+ RSS news sources with AI categorization, sentiment analysis, entity recognition, and geotagging
- 40+ geospatial layers including active conflicts, ADS-B aircraft, AIS vessels, 220+ military bases across 9 countries, nuclear facilities, NASA satellite fire detection, submarine cables, internet outages, power grids, protests, social-media hotspots, sanctions, and elections
- ADS-B explained: Automatic Dependent Surveillance–Broadcast transmits GPS position, altitude, speed, callsign, and heading on 1090 MHz every ~0.5 seconds. Civilian aircraft are required to broadcast; many military transport, tanker, and training flights also broadcast, which is why the platform can display some military activity.
- AIS explained: Automatic Identification System requires vessels over 300 tons to broadcast identity (MMSI, name), position, course, speed, destination, and cargo type on VHF (~162 MHz). Ships that go silent — "going dark" — are flagged because silence can indicate illegal fishing, sanctions evasion, or covert operations.
- AI features: Daily/weekly briefings, anomaly detection, and "convergence alerts" when multiple weak signals (e.g., military flights + protests + internet outage) co-occur. Local LLM inference is supported through Ollama or LM Studio for offline, private, key-free operation.
- Country Instability Index (CII): 0–100 score per country derived from military activity, protest intensity, news sentiment, and structural risk factors. Intended as a thermometer, not a forecast.
- Tech stack:
- Frontend: deck.gl + MapLibre GL for WebGL-rendered 3D globe (rotation, zoom, tilt, multiple projections, 3D terrain overlays)
- Backend: Vercel Edge Functions hosting 60+ API endpoints, Railway relay for WebSocket realtime updates, edge caching for low latency and pay-per-request scaling
- Desktop: Tauri (TypeScript frontend, Rust backend), cross-platform installers under ~10 MB, lower memory footprint than Electron
- Use cases: citizens tracking global events, investors monitoring supply-chain risk (e.g., Suez Canal), journalists hunting leads from unusual military or ship movements, and developers self-hosting or extending the platform.
- Ethics and limitations: ADS-B and AIS data can be spoofed; social-media clips can be recycled out of context; AI summaries may misinterpret nuance. The documented best practice is cross-verification across independent sources. Privacy concerns around large-scale aggregation of public broadcasts, and information-overload anxiety, are explicitly acknowledged — Happy Monitor is the project's response to the latter.
- Getting started:
- Web: open https://worldmonitor.app (no signup)
- PWA: install via Chrome/Edge install icon or Safari "Add to Home Screen" for offline tiles and push notifications
- Self-host:
git clone https://github.com/koala73/worldmonitor.git, thennpm install, copy.env.exampleto.env, andnpm run dev - Desktop: download installers from https://github.com/koala73/worldmonitor/releases
- Outlook: anticipated directions include satellite imagery, IoT, and blockchain data integration; predictive analytics; collaborative analyst workflows; and VR/AR visualization — paired with cautions about data flooding, disinformation poisoning, and an OSINT capability gap.
- Repository: https://github.com/koala73/worldmonitor
- Web app: https://worldmonitor.app
- Releases (desktop): https://github.com/koala73/worldmonitor/releases
- License: AGPL-3.0