OpenRescue is a battle-ready, fully open-source emergency coordination platform designed to deliver faster, smarter, and more reliable incident response — even when the internet fails. Built from the ground up for privacy, resilience, and real-world impact, OpenRescue turns ordinary phones and local networks into a coordinated emergency response system that an institution can self-host and own. It’s not a prototype — it’s a practical, deployable solution engineered to save time, improve accountability, and scale from a single campus to a whole district without a single proprietary API key.

What it does (short)

OpenRescue lets anyone report an emergency in one tap (optionally anonymous), broadcasts the alert to local responders in real time (or over LAN when offline), auto-suggests the best responder using a workload- and distance-aware algorithm, and creates an auditable incident trail with heatmaps and risk analytics — all using open-source components and self-hosted infrastructure.

Key features (concise, powerful)

• Offline-first reporting: submit incidents with no internet; alerts queue locally (IndexedDB/SQLite) and auto-sync when connectivity returns.

• LAN/WebRTC broadcasting: alerts propagate over local networks so nearby responders see incidents even during network outages.

• Real-time dashboard: WebSocket-driven responder view with live markers, status updates (Pending → Assigned → Resolved), and quick assignment actions.

• Smart responder allocation: algorithmic suggestions that consider distance, current workload, and availability to minimize response time.

• Privacy-first design: optional anonymous reporting, role-based data visibility, and self-hosted storage (Postgres) — no third-party data sharing.

• Risk analytics & heatmaps: cluster incidents, detect hotspots, and surface time-based trends to move from reactive to preventive safety measures.

• Self-hostable & FOSS: Dockerized deployment, MIT/GPL-compatible licensing, and full code visibility so institutions control their data and cost.

Technical highlights (novelty & engineering strength)

• Resilient sync engine: deterministic conflict resolution and idempotent sync ensure consistent incident state across intermittently connected devices.

• Local ML prioritization (optional): small, explainable models (scikit-learn) run entirely on-prem to surface high-priority incidents without cloud ML.

• Efficient network discovery: lightweight peer discovery over LAN for rapid device pairing—no central server required for local alert propagation.

• Open mapping stack: OpenStreetMap tiles served from a local cache so map features work offline and avoid proprietary mapping locks.

• End-to-end developer ergonomics: well-documented REST + WebSocket APIs, modular microservices (or a monolith option), and a clear test suite for rapid iteration.

Impact & who benefits

• Universities and colleges: campus security can respond faster and identify unsafe hotspots.

• Small towns & rural clinics: institutions without reliable internet gain robust local emergency coordination.

• Event organizers: offline-capable alerts during large gatherings reduce chaos and improve responder routing.

• NGOs & disaster relief: deployable into temporary infrastructure where connectivity is unreliable.