DroneMonitor | Passive Drone Detection as Edge AI & SaaS
Industry: Security, critical infrastructure, manufacturing | Period: 2026-present | Role: IP7 Consulting | Status: MVP phase
Starting Point
Owners and operators of industrial plants, data centres, research and logistics sites and high-value private properties currently have no simple way to determine whether a drone is flying over their premises, how long it stays, whether it broadcasts a Remote ID and what reliable evidence is available after an incident. Professional counter-UAS systems are expensive, complex and designed for critical infrastructure | not for small and medium-sized sites.
Product Promise
DroneMonitor detects suspicious aerial objects over a defined site, alerts the responsible people and generates a traceable event package | without ever actively interfering with the flight.
Solution Architecture
- Edge appliance (NVIDIA Jetson Orin Nano class) with encrypted local storage, Docker runtime and store-and-forward buffer
- One or more RTSP / 4K IP cameras with configurable detection zones and privacy masks
- Optional Remote ID receiver via Bluetooth and supported Wi-Fi methods
- Local computer vision pipeline: GStreamer / OpenCV, ONNX Runtime or TensorRT inference, object tracking across frames, confidence scoring
- Sensor fusion of visual tracks and Remote ID beacons into a single event with clear source labelling
- Multi-tenant SaaS back-end (ASP.NET Core, PostgreSQL, S3-compatible object storage, OpenID Connect) with device registry, event and media service, alert dispatch and audit log
- React / TypeScript dashboard with event triage, evidence export (ZIP + report), device and site health, role management
Core Principles
- Passive: detection only, no jamming, no takeover, no active countermeasures
- Edge first: analysis on site | images leave the site only as metadata
- Privacy by design: privacy masks, retention rules, no facial recognition
- Fail safe: continued local operation during internet outages, store-and-forward
- Secure by default: device identity, signed updates, mTLS
- Explainable events: complete metadata per event
Explicitly Out of Scope
Radar, acoustic sensors, active countermeasures, GNSS or Wi-Fi jamming, spoofing, drone takeover, facial recognition, automatic identification of private individuals and continuous cloud upload of all video streams.
Targets (Concept / MVP)
Business Model
Hardware sale or rental, one-off site survey and installation, tiered SaaS subscriptions (Monitor / Verify / Evidence / Managed Site) and optional integration into existing security and building management systems.
Technologies Used
NVIDIA Jetson Orin, GStreamer, OpenCV, ONNX Runtime, TensorRT, Python, C++, ASP.NET Core, PostgreSQL, S3-compatible object storage, OpenID Connect, React, TypeScript, Docker, mTLS, MQTT, Bluetooth LE, Remote ID (ASTM F3411 / EU Delegated Regulation 2019/945).
Status: Concept and MVP preparation | not yet in production. This entry documents the product design; a paid pilot phase is being agreed with initial customers.
Interested in a pilot installation?
If you are responsible for a site and would like to evaluate passive drone detection early, we are happy to discuss a pilot agreement.