Visual-Inertial Navigation System
Localization in any environment
VINS provides accurate global positions in any environment. Especially for GNSS denied environments localization is a challenge. State-of-the-art performance in these environments is achieved by fusing RTK GNSS, camera, IMU, and other sensor modalities. This ensures reliable and accurate localization required for autonomous driving and mobile robotic applications.
Going beyond localization
VINS goes beyond localization by outputting position (Latitude, Longitude, Altitude) and pose (6 degrees of freedom). Initialization is instantaneous, so the pose output can be used directly after the system start. Instant pose availability is crucial for autonomous systems and mobile robots.
GNSS safety concept: reliable by design
Providing accurate signal health metrics is mandatory for safety-critical applications. VINS continuously outputs signal health in the form of Protection Level (PL) for both position and pose information (accuracy and integrity). Protection Level is an estimate of maximum possible error in position. The proprietary multi-modal sensor fusion evaluates integrity of the measurements from different sensors and provides higher reliability and redundancy of the whole system.
The system is designed to be ASIL-B certification ready. With an ASIL VINS a high integrity position can be provided even during GNSS outages or IMU drift. Integrity estimation ensures reliability of output solution for safety applications.
VINS is natively compatible with ROS, the standard framework for building autonomous systems. A driver and RosNode make VINS a drop-in solution: pose, position, and the image stream are be published to the respective topics.
Accurate global positioning (also in GNSS denied environments)
Reliable signal health, covariance matrix
Direct SLAM technique: instead of using features, it directly operates on image intensities
Automotive grade & ASIL-B ready (coming soon)
Map based relocalization (coming soon)
Global positioning: 0.6cm*
Hold-over: 0.2% of distance travelled*
Output rate: 1hz - 200hz
Formats: NMEA0183, ASCII, Binary
Native ROS Support
Ground truth is notoriously hard to establish, yet crucial for testing of autonomous systems as well as location services. Artisense VINS beats conventional ground truth systems where such systems don’t work: tunnels, mountains or urban centers.
Many AI-based smart & big data apps rely on images as input in order to extract intelligence. Without precise geo-referencing of pictures, they lack geo-spatial and temporal information which unlocks the true value in data analytics. Artisense VINS delivers all in one system.
Map makers are dependent on accurate localization to register maps in globally standardized coordinate systems. Any error can lead to fatal consequences when referenced by an AV or in combination with other maps. Artisense VINS provides ground truth quality localization.
Automated Guided Vehicle
Early use of autonomous technology is found in so called AGVs, which travel along pre-defined trajectories in geofenced areas. Moving along such “virtual rails” and understanding the environment is enabled by Artisense VINS far more robustly than with high-end INS.
Robust localization for mobile robots traversing outdoor, indoor or both environments. Artisense VINS meets the requirements towards small form factor, low compute and low TCO for mobile robotic applications in e.g. warehouses and manufacturing or delivery applications.
Cities, TelCos, Transport, Real-Estate and Infrastructure managers all have many assets deployed in public space. The inventorization and maintenance of such assets is a labor intense task lacking automated detection and robust localization. Artisense VINS enables both.