Analysis, technical deep-dives, and situation reports from the OpenWar Intelligence project. Open source defense analytics for the OSINT community.
How we built a real-time military intelligence platform that tracks 117+ aircraft, 49+ ships, and 15+ satellites across 27 nations — all in a single HTML file with zero dependencies. The architecture decisions, data fusion approach, and neural network prediction engine behind version 5.0.
A technical deep-dive into how OpenWar Intelligence maps A2/AD systems, GPS jamming zones, and radar emissions across 9 frequency bands.
How our nuclear domain monitor tracks the world's nuclear-armed nations — patrol zones, treaty compliance, and real-time posture assessment.
Analyzing the strategic chokepoints from Hormuz to Malacca — blockade risk modeling, naval force deployments, and economic impact scoring.
The 12-24-1 architecture trained on 37 historical conflict events. Online learning, 60-second update cycles, and predictive timelines from 1h to 24h.
See all these intelligence modules working together in real-time. Single HTML file, no installation, no API keys required.
Get Access — Free DownloadOpenWar Intelligence is now live. Version 5.0 represents over 4,400 lines of pure JavaScript and CSS packed into a single HTML file that delivers real-time multi-domain military situational awareness — no server, no API keys, no build tools, no dependencies.
This post covers why we built it, how the architecture works, and where the project is headed.
Military situational awareness has traditionally required either classified access, expensive commercial platforms (Janes, Palantir, BAE Systems), or fragmented open-source data scattered across dozens of disconnected tools. An analyst wanting to correlate aircraft movements with naval deployments, satellite passes, SIGINT activity, and economic warfare signals had to manually cross-reference multiple interfaces.
We asked a simple question: what if all of that lived in one browser window?
OpenWar Intelligence was built on three principles: everything runs client-side, every data source is open and publicly accessible, and the entire platform ships as a single file anyone can download, audit, and modify.
The decision to build a 265KB single-file application was deliberate. No webpack, no React, no npm install. The entire platform is vanilla JavaScript with inline CSS and SVG assets. This design choice delivers several operational advantages:
The application initializes by loading a Leaflet.js map canvas, then sequentially activates 13 intelligence modules. Each module operates independently with its own data pipeline and update cycle, but feeds into a central fusion engine that detects cross-domain signal convergence.
The fusion alert engine correlates 9 signal types: AIR, SHIP, ML-PRED, AIR-BRIDGE, A2AD, THERMAL, NAVAL, CYBER, and NUCLEAR. When multiple signal types converge within a defined geographic radius and time window, the engine generates fusion alerts at three severity levels — Elevated, High, and Critical — displayed as sonar-pulse animated markers on the operational map.
This is the core innovation: not any single data source, but the correlation across domains. An aircraft surge near a thermal hotspot detected by NASA FIRMS, combined with naval repositioning and a neural network escalation prediction, tells a much richer story than any signal alone.
One of the most technically ambitious features is a pure JavaScript neural network with a 12-24-1 architecture. No TensorFlow, no ONNX runtime — just matrix math in vanilla JS running in the browser's main thread.
The network was trained on 37 historical conflict events using features extracted from publicly available pre-conflict indicators: military mobilization patterns, diplomatic signal changes, economic sanctions timing, naval deployment shifts, and airspace activity anomalies.
The model performs online learning — it retrains every 60 seconds against the current data feed, adapting its weights to evolving situations. The output is a predictive timeline generating escalation probability scores at 1h, 6h, 12h, and 24h horizons, with trajectory estimation on the map.
The goal isn't to predict wars — it's to surface patterns that human analysts can evaluate with their own judgment. The neural network is a signal amplifier, not an oracle.
Version 5.0 ships with a comprehensive multi-domain intelligence suite:
Tracks 117+ military aircraft across 27 nations using multi-source data fusion from FlightRadar24, OpenSky Network, and ADS-B Exchange. Includes model identification with SVG silhouettes and technical specifications for 50+ airframes — fighters, bombers, tankers, ISR platforms, and transport aircraft.
Monitors 49+ naval vessels including carriers, destroyers, cruisers, frigates, and submarines. AIS integration provides real-time positions with vessel classification and threat assessment. The chokepoint module covers 9 global maritime bottlenecks with blockade risk scoring, naval force deployment tracking, and incident monitoring.
15+ military and reconnaissance satellites tracked in real-time with pass prediction. Sentinel-2, Copernicus, and NASA FIRMS integration for satellite imagery analysis and thermal intelligence. Multi-layer tile switching between dark operational map, satellite imagery, and terrain views.
Full nuclear posture assessment across 9 nuclear-armed nations. Tracks SSBN patrol zones (10), ICBM silos (12), nuclear-capable bombers (8), C3 command nodes (12), and treaty compliance status (6). Includes deployed/reserve/retired inventory counts and nuclear doctrine analysis for each state.
Maps the electromagnetic battlespace: A2/AD systems (29), GPS jamming zones (7), electronic warfare platforms (10), radar emissions (9), and cyber warfare operations (7). Real-time EW spectrum monitoring across 9 frequency bands from VLF to EHF.
Tracks 8 active sanctions regimes with SWIFT/asset freeze status, 9 strategic commodities with war impact analysis, and 8 critical supply chains with vulnerability scoring. A weighted economic threat index provides a composite indicator of economic warfare intensity.
19 RSS feeds from defense and geopolitical sources (Reuters, AP, Jane's, The War Zone, Breaking Defense, TASS) power a real-time news ticker. NASA FIRMS VIIRS thermal hotspot data detects fires, explosions, and artillery activity with direct links to Sentinel Hub, Copernicus Browser, and Planet Labs imagery.
OpenWar Intelligence is released under the MIT License. This isn't just a licensing choice — it reflects a conviction that situational awareness should not be a privilege.
Commercial C4ISR platforms cost hundreds of thousands of dollars and are available only to governments and large defense contractors. Journalists covering conflict zones, NGO workers assessing humanitarian crises, independent researchers studying escalation dynamics, students learning intelligence analysis — none of these communities have had access to integrated, multi-domain situational awareness tools.
The open-source model also enables community contribution. The codebase is structured so that new intelligence modules can be added by anyone with JavaScript knowledge. Want to add a cyber threat intelligence feed? A weather overlay for military operations? A logistics tracking module? Fork the repo, add your module, submit a pull request.
Version 5.0 is the foundation. The roadmap includes:
OpenWar Intelligence v5.0 is live at openwarintelligence.org. The full source code is available on GitHub. Download it, run it, break it, improve it.
If you find it useful, star the repository. If you find a bug, open an issue. If you build something cool on top of it, let us know.