DroneWatch AI: Building a Multi-Sensor Drone Detection Demonstrator

Description and use

DroneWatch AI demonstrates that fusing multiple weak sensor streams can reduce false alarms and provide clear threat assessments without additional hardware.

Problem

Commercial drones threaten critical infrastructure, but specialized counter‑drone hardware is expensive and coverage is sparse. Single sensors produce noisy false alarms that overwhelm operators.

Outcome

The demonstrator shows that software fusion of existing sensors can reduce false alarms and provide clear threat assessments without additional hardware.

Workflow

• Collect live ADS‑B and Remote ID feeds
• Ingest radio‑frequency, acoustic, telemetry, registry, and human reports
• Apply deterministic rules to filter authorised flights
• Use AI layer to explain threat level
• Push scored track to existing command interfaces (ASTM F3411, ADS‑B, ASTERIX, U‑space/UTM, STANAG 4586/4609, Link 16)

Evidence

The drone-threat signals are simulated. Two public feeds are live and real: ADS‑B air traffic and, since 25 September 2026, network Remote ID

Source

A HIGH assessment requires at least two independent sensors to agree and a contact with no matching authorised flight

Source

It stays strictly defensive: detection and early‑warning only, with any response left to a separate, authorised, human‑in‑the‑loop system

Source