How will GNSS developments impact AV evolution?


Manuel Del Castillo explores how GNSS software enhancements in the chip will allow the accuracy required to be part of Level 4 driving systems

It was during the Sputnik era that scientists were able to track satellites by detecting shifts in radio signal waves known as the Doppler Effect, which led to the development of GPS. In the mid 1960s, the US Navy conducted satellite navigation experiments to track nuclear missile-loaded submarines. Six satellites orbiting the poles enabled submarines to detect changes in satellite Doppler within a few minutes, allowing them to pinpoint their location.

Early in the 1970s, the Department of Defense (DoD) sought a robust, stable satellite navigation system. Embracing previous ideas from Navy scientists, the DoD decided to use satellites to support their proposed navigation system. In 1978, the Department of Defense launched the first Navigation System with Timing and Ranging (NAVSTAR) satellite. The 24-satellite system became fully operational in 1993.

As of today, GPS is a multi-purpose, space-based radio navigation system operated by the United States Air Force for national defense, homeland security, civil, commercial, and scientific purposes. GPS currently provides two levels of service: Standard Positioning Service (SPS) which uses the coarse acquisition (C/A) code on the L1 frequency. Due to its limited performance and challenges with availability, GPS was not adopted as a primary localisation sensor in early prototypes. In that era, however, GPS was also in its infancy. Un-degraded GPS was first used by civilians in 2000. In the years since then, both autonomous driving and GNSS have matured into commercial products. Level 2 systems are available for highway driver assistance, while Level 4 systems are nearing readiness for ride-sharing and city deliveries. GNSS receivers are now deployed in billions of consumer and automotive devices, bringing navigation to the masses and driving down costs.

By 2020, satellite navigation included four independently operated GNSS constellations: the US GPS, Russian GLONASS, European Galileo, and Chinese BeiDou, leading to a substantial boost in availability. Aside from the increased capability in signal acquisition, atmospheric correction, precise positioning, multipath mitigation, and spectrum diversity for resilience to radio interference due to satellite modernisation, multiple civil frequencies are now transmitting modern signals. These recent enhancements have led to significant industry investment in continent-scale GNSS monitoring networks, which now deliver GNSS corrections at scale for precise positioning in the automotive market.

S-GNSS improves the accuracy and security of GPS technologies for the automotive industry

How is GNSS used in Level 2 ADAS and beyond?

For Level 2 autonomous driving systems, lane-level GNSS localisation and HD maps can unlock the complex manoeuvres required to safely and confidently move beyond lane keeping. By using high precision GNSS, autonomous driving systems are able to monitor perception system outputs, provide high assurances of their accuracy, and anticipate upcoming road elements in order to manoeuvre into the appropriate lanes for interchanges, exits, and merges.

This challenge is well within the reach of GNSS technology. With lane-determination accuracy greater than 95% on US freeways, production-ready ASIL-rated GNSS chipsets supporting multiple constellations and multiple frequencies are linked to continent-scale correction services. Providing oversight and becoming a primary sensor for lane-level manoeuvres requires safety-of-life risk management. GNSS is capable of binding fault risks beyond three meters per hour to better than 10-7 per hour.

For Level 4 autonomous driving systems, GNSS has demonstrated the accuracy required to complement LiDAR-based localisation, providing critical information when LiDAR systems experience outages from weather, observability difficulties in sparse environments with limited distinctive 3D structures, or hardware faults. The microwave signals of GNSS, on the other hand, are not affected by rain, snow, or fog. In many ways, GNSS and LiDAR are ideal complementary sensors—GNSS works particularly well in open-sky environments with few features, while LiDAR works particularly well in environments filled with geometric features.

The only source of absolute location accuracy

GNSS based positioning is the only globally available technology that can provide a reliable, high-precision source of absolute position, ie not referenced to anything else. GNSS signals, however, are transmitted with the power of a 45-watt lightbulb, and from 20,000km away, so any obstruction can make it even harder for a receiver to pick them up. Buildings can block them completely, just like the glow from a dim light can be blocked by a blackout curtain. With foliage they can still pass through, but the effect is like a thin curtain making a dim light even dimmer.

This means that automotive manufacturers need to overcome the big issue of multipath, before they can fully utilise all the benefits associated with GNSS technologies. Multipath can occur in any environment whereby obstacles such as buildings, trees, even mountains can ‘bounce’ GNSS signals before reaching the receiver, creating multiple signal paths. This can cause errors in the GNSS position calculation and can be particularly problematic for AVs, where accurate positioning is critical for safe and reliable operation.

The future of accurate GNSS for ADAS and autonomous vehicles

Looking forward, GNSS software enhancements in the chip can solve multipath and when combined with RTK/PPP corrections, will allow the accuracy required to be part of Level 4 driving systems, all within a high reliability system with safety guarantees. A true driverless vehicle can be safely operated with precision GNSS as a foundational pillar of absolute location that opens up accessibility and availability in challenging environments.

In the future, autonomous vehicles and GNSS will continue to evolve together. For example, researchers are examining the possibility of an end-to-end machine learning approach to self-driving. As with current implementations, GNSS can enable these architectures because it enables robustness and redundancy. Additionally, GNSS is constantly improving and evolving. Ground monitoring infrastructure, new integrity algorithms, and user equipment are adding new capabilities. Globally, the industry is moving towards ground station networks and LEO constellations providing high availability and high accuracy with safety-of-life integrity guarantees.

Autonomous driving relies heavily on safety guarantees to ensure the reliability and trustworthiness of the systems that control these vehicles. With high accuracy and integrity, GNSS systems can help prevent accidents and malfunctions, providing the necessary confidence for widespread adoption. Robust safety guarantees build public trust and pave the way for a cost-effective future of autonomous transportation.


About the author: Manuel Del Castillo is Vice President of Business Development at FocalPoint



Source link

About The Author