VANAV GNSS independent visual navigation
GNSS-Independent Visual Navigation

Reliable Positioning with Visual Navigation

VANAV derives position, velocity and orientation from onboard camera imagery, supporting mission continuity in jamming and spoofing environments.

Explore field evidence
Passive Navigation Active
Position outputPosition (2D/3D), velocity, acceleration, orientation, confidence score
Confidence score2D ~7 m @ 1500 ft; 3D ~10 m @ 1500 ft
Mission continuityReturn-to-Home (RTH) and route continuity
10 Hz 10 Hz (instantaneous output)
25 Hz 25 Hz (platform/conditions dependent)
50-900 m 50-900 m
~40 W ~40 W
Operational Logic

A positioning line independent of external signals

VANAV runs visual odometry, visual localization and PnP-based methods on-edge, providing alternative position information to the flight computer when GNSS signals weaken or are manipulated.

Passive navigation in GNSS-denied environments
AI-powered visual localization
Architecture resistant to jamming/spoofing
Low-latency real-time processing
Return-to-Home (RTH) and route continuity
Lightweight, compact, low power
Visual Positioning

Terrain imagery becomes navigation data.

Camera streams generate meaningful position output for the flight computer through terrain matching, visual odometry and confidence scoring.

VANAV visual terrain mapping
Visual mapping
VANAV terrain matching and visual localization
Terrain matching
Field Footage

Test results, search scenario and product story on one surface

Field Tests - Summary
VANAV Teaser
VANAV Search
Technical Values

Technical Specs

Data Output Rate 10 Hz (instantaneous output)
Image Processing 25 Hz (platform/conditions dependent)
Operating Temperature -20 °C ~ +55 °C
Operating Altitude Range 50-900 m
Power (Field) 15-45 W
Generated Data Position, velocity, orientation, acceleration, confidence score

* Typical values from PDF; field ranges provided in the technical table.

VANAV passive image processing over terrain
Passive image processing
VANAV route continuity with visual navigation
Route continuity
Architecture Flow

VANAV is a modular system integrated with the flight computer, combining visual odometry, visual localization, and PnP-based methods. It runs on-edge without external connectivity.

  • Independent module integration
  • Bi-directional communication with mission computer
  • Activation without interfering with manual control
  • Operation in jammer/spoofing environments
Field Tests

Field Tests - Summary

Kalecik/TR - 12.11.2022: Initial flight tests.

HAVELSAN - Karaağlı - 01.03.2023, 04.07.2023 (noon), 12.07.2023 (evening): Multiple mission profiles.

Figures compiled from field summaries.
Test-1 (Mar’23) 500 m, 30 min, 837 images → 811 coordinates | Accuracy: ~20 m (2D), ~37 m (3D)
Test-2 (Jul’23/Noon) 500 m, 900 images → 751 coordinates | Accuracy: ~24 m (2D), ~41 m (3D)
Test-3 (Jul’23/Evening) 500 m, 2000 images → 1751 coordinates | Accuracy: ~22 m (2D), ~36 m (3D)
Category Visual Navigation, Computer Vision, Artificial Intelligence
Scope Passive position, velocity, and orientation in GNSS jamming/spoofing environments
Platform Fixed/rotary-wing UAVs and various drone types
Operating System Linux (Ubuntu)
Processing Unit NVIDIA Jetson AGX Orin Industrial 64 GB
Outputs Position (2D/3D), velocity, acceleration, orientation, confidence score
Accuracy (Typical) 2D ~7 m @ 1500 ft; 3D ~10 m @ 1500 ft
Power Consumption (Typical) ~40 W