High accuracy positioning for navigation, ADAS and autonomous driving

Lane level navigation & ADAS paving the way to the autonomous vehicle

Ahead of the deployment of autonomous vehicles, there is a new generation of navigation experiences that will delight drivers. One of these is the ability of GPS navigation systems to help drivers switch lane, in a secure manner, based on their exact lane position and their destination.

This feature will also envision the HMI presented to the car “driver“ and passengers when the vehicle is automated in Level 3 or Level 4 (SEA autonomous car level). This HMI will help the passenger to anticipate the next maneuver of the car and hence decrease the anxiety of self-driving mode.

But a pre-requisite for lane level navigation is to have the exact position of the car, something that cannot be provided accurately by standard GNSS technology today.

The same precise positioning is required for ADAS (Advanced Driver Assistance Systems) functionalities such as Lane centering, lane departure warning, lane change assistance , Intelligent Speed Assistance (ISA).

Furthermore, other location-based services such ride hailing, car-sharing, real-time parking information, etc. will take advantage of a significantly improved positioning of the car, its driver and passengers.

High Accuracy GNSS and sensor fusion

Developers of ADAS solutions have traditionally used cameras, infrared cameras, ultrasonic sensors, radar and Lidar sensors to position the vehicle on its lane with precision. GNSS technologies (GPS, Galileo, BeiDou, Glonass) have generally been either neglected or completely dismissed because of their (perceived) lack of high precision.

However, with Geoflex’s GNSS correction services a large part of GNSS errors are eliminated, allowing decimeter or centimeter positioning accuracy, based on the use case requirements, in a lean hardware setup.

Furthermore, thanks to Geoflex sophisticated positioning engine, GNSS measurements can be easily fused with a number of other sensors embedded in the average car (odometer, car heading based on steering wheel angle, inertial sensors), or embedded in automated vehicles (radar, lidar, mono and stereo cameras). This sensor fusion improves the positioning accuracy and its robustness, mitigating situations where GNSS signal is not available or in limited supply with few satellites in view (typically under a bridge or in urban canyons between high-rise buildings).

A system using Geoflex GNSS corrections coupled with an HD map will provide extra safety for driving an autonomous vehicle (L3 to L5) primarily equipped with lidar and radar sensors, it will provide longer range of vision for the car.

Hypergeolocation Everywhere

The great advantage of GNSS, compared to other ADAS sensors, lies in providing an absolute position independently of external landmarks (such as lane marking, surrounding infrastructure, etc…) and whatever the weather and light conditions are (snow, rain, night).

Unlike a number of other high accuracy GNSS systems, Geoflex’s service, available worldwide since 2018, is free of any nearby base stations. This represents a great advantage for global car makers and automotive suppliers: wherever a Geoflex-enabled system is sold it just works.

Geoflex correction data feed is provided under a standard industry format (RTCM), working with all GNSS receivers and offering data for all constellations: GPS, Galileo, BeiDou and Glonass.