Vision-based navigation system productline


The enhancement of the A-ROX system

With computer vision sensors including AI algorithmus


Position accuracy

0.01 km/h






The V-ROX system is the enhancement of the A-ROX GNSS-INS positioning system. It includes 3 Multi-frequency, Multi-GNSS (GPS + Galileo + Glonass + Beidou) receivers, FOG-performance MEMS IMU, CAN interfaces (for reception of vehicle data (wheel odometry and steering angle) and solution output), 5G module for reception of RTK/ PPP corrections (e.g. from ANavS® G-ROX reference station, and the powerful ANavS® sensor fusion on a single board. The latter one performs a tight coupling of all sensor data with an Extended Kalman Filter (EKF).

On top of the standard sensors a fully integrated computer vision module is equipped, that can be flexibly configured with cameras and 3D-LiDAR.

The default camera configuration is a GMSL2 global shutter wideangle stereo-camera with high-performance IMU, external synchronization and integrated depth estimation, localization and object detection features.

The default LiDAR configuration includes a high-quality 128-channel 3D-LiDAR.

On the one hand, both the camera sensors and the LiDAR sensor provide an additional odometry or positioning information through visual-inertial odometry or Simultaneous Localization and Mapping (SLAM). On the other hand, 2D and 3D maps of the environment or road can be generated. Furthermore, object-detection with 2D or 3D bounding boxes, enabled by deep learning algorithms, allows to add semantic information to the maps and further enhances the SLAM performance.

The additional positioning information obtained from the computer vision module is coupled with GNSS, IMU and wheel odometry information in the tightly coupled sensor fusion, also called centralized filter. This leads to an improved positioning accuracy to overcome also the most challenging environments.

The V-ROX supports both Real-Time Kinematics (RTK) (verlinken) and Precise Point Positioning (PPP) (verlinken). RTK positioning uses measurements from a terrestrial reference station (e.g. from ANavS® G-ROX reference station, that are sent to the V-ROX via internet. PPP uses satellite position, clock offset, phase and code bias corrections that are sent via internet or directly via satellite (using Galileo’s High Accuracy Service, HAS). The RTK positioning typically provides a higher positioning accuracy and faster convergence than PPP, but requires at least a partial internet coverage of your target application region and proper reference station infrastructure.

The V-ROX is ready to use out-of-the-box. With its vacuum cup it can be fast mounted for any kind of automotive test run.

3D LiDAR SLAM Performance

Overview V-ROX

Two different versions of V-ROX are available:

  • Autonomous vehicle testing
  • Survey and mapping
  • Georeferencing of cameras and LiDAR sensors
  • Logging of parameters during vehicle tests and machine automation.

The V-ROX system comes with an integrated 5G module, providing access to RTK and PPP correction data and enabling remote view and system configuration very user-friendly.

  •  Gigabit Ethernet
  • WiFi
  • USB-C and up to 4 CAN -FD channels
  • Trigger events, like wheel ticks, can be fed through the GPIOs into the sensor fusion framework.

There are several output formats available with a maximum rate of 200 Hz (1000 Hz optional): ROS 2, NMEA, ACOM and CAN.

The patented tightly coupled sensor fusion algorithms integrate survey-grade GNSS raw measurement data (multi-constellation, multi-frequency), RTK and PPP (HAS and terrestrial) correction data, inertial sensor data (FOG-grade MEMS IMU), odometry data, LiDAR data and camera data for a highly accurate position, velocity and attitude solution, even in challenging GNSS environments. AI based algorithm supports the localization solution for a unique positioning performance.

A powerful forward-backward post-processing engine is available to further enhance the accuracy of the position, velocity and attitude solution with all recorded sensor raw data.

The V-ROX system is built on a new modular hardware platform, delivering improved processing capabilities and upgraded interfaces. The system is configured with three GNSS antennas. The processing units, including two embedded GPU units, are part of the V-ROX system. A powerful configuration and visualization software is implemented as a web app, easily accessible from different kinds of devices, including laptops and tablets.

It is directly hosted on the system with no need for installation of software on your device. On top of the local app, a
cloud-based fleet management feature is available for controlling all your V-ROX devices and accessing all sensor data from anywhere.

Our features - Your advantage

Highly Accurate Positioning


AI Inside

RTK-PPP Handover


HAS + OSNMA Technology

Web App

Easy System Integration


ACOM Data Stream


Minimal Warm-up time

Made in Germany

Quick overview



Equipped for every situation with the need of high-precise positioning and navigation

R&D Projects

Our R&D team is constantly working on researching, developing and implementing new technologies to master the challenges of tomorrow.


Our publications provide well-founded insights, present innovative solutions and set trends.

Any questions?

Feel free to ask!