DominIQ
A concurrent MAVLink mission controller coordinating four autonomous assets, with visual target localization and a Meta Quest 3 interface. Built at Hack The North 2026.
WINNER / HACK THE NORTH 2026
Two robots converging on a target
Illustrative demoScout A is searching. Scout B is standing by.
An illustrative coordination sequence, not a recording of DominIQ.
- Autonomous assets coordinated
- 4
- Vessel detection accuracy
- 98%
About this project
Coordinated four heterogeneous autonomous assets through a concurrent MAVLink mission controller with dynamic search, interception, target confirmation, tracking, and lost-target recovery.
Achieved 98% vessel detection accuracy by fine-tuning YOLO26m and fusing camera detections with robot altitude, orientation, and camera geometry to localize targets in a shared coordinate system without relying on GPS.
Built a Meta Quest 3 interface using WebXR and Meta IWSDK to display telemetry, camera feeds, and targets in 3D. DominIQ was a Hack The North 2026 winner.
How it works
- ASSET TELEMETRYPose + camera feeds
- TARGET DETECTIONYOLO + shared coordinates
- MISSION CONTROLConcurrent MAVLink missions
- QUEST INTERFACEWebXR + Meta IWSDK
Built with
- Python
- PyMAVLink
- YOLOv8
- ArduPilot
- OpenCV
- WebXR
- TypeScript