Just add AI for smarter satnav... from rogue drone detection to safer driving
Last Updated: 07/08/2023 06:10 Created at: 04/08/2023 06:43
In an era where data is generated at an unprecedented rate, Artificial Intelligence (AI) emerges as a powerful tool to harness and utilize this wealth of information. From space exploration to everyday terrestrial applications, AI has become an integral part of modern advancements. ESA’s NAVISP programme supports companies and research institutions to push the boundaries by applying AI to PNT technologies. The results are a growing portfolio of prototype services, ranging from improved space and Earth weather forecasting to safer autonomous driving and even the identification of rogue drones in sensitive airspace.
The use of PNT information has become ubiquitous in our lives, extending its utility far beyond transportation. From precision farming to power grid management, financial trading to communication links, the applications are diverse and indispensable. The key to achieving this lies in merging Global Navigation Satellite Systems (GNSS) with other cutting-edge technologies, including inertial sensors, machine vision, and the emerging 5G and 6G networks. However, it is the integration of AI that brings a game-changing aspect to this revolutionary endeavour. Rafael Lucas Rodriguez, Head of the NAVISP Technical Programme Office, highlights AI's potential in extracting meaningful information and uncovering hidden patterns within vast datasets – a feature particularly useful in satellite navigation.
One exemplary project, CAMALIOT, leveraged an Android app to collect raw GNSS data from smartphones owned by over 12,000 volunteers. This immense dataset, consisting of more than 131 billion measurements covering Europe and the entire globe, has been employed in machine learning algorithms to identify patterns related to space weather and tropospheric effects, significantly enhancing weather forecasting accuracy.
In the realm of autonomous driving, the AIGNSS project sought to optimize GNSS positioning to ensure safer mobility. The inherent variability of GNSS performance due to environmental factors like signal reflectance and limited satellite visibility in dense urban settings is a significant challenge. AIGNSS overcame this by analysing real-world driving data using multiple satnav constellations, including Europe's Galileo, and employing AI algorithms. While the improvements achieved were not drastic, the project laid the groundwork for more comprehensive algorithm development, which is essential for future advancements in autonomous vehicle safety.
The potential of AI in maritime navigation is exemplified by the Trondheimsfjorden Autonomous Vessel Test Area in Norway, established with NAVISP support. This first-ever state-of-the-art test zone is equipped with advanced technologies such as communication infrastructure, positioning sensors, and Internet of Things connectivity. AI is deployed to analyse test outcomes, ensure optimal performance, and guide autonomous vessels efficiently.
Finally, as the integration of AI in the PNT domain progresses, it is likely to revolutionize the way we navigate and interact with technology. The success of these initiatives signifies that AI has become an indispensable tool in extracting valuable insights from vast datasets, leading to safer and more efficient navigation across various domains. Moving forward, continued research, collaboration, and innovation in AI will be vital in unlocking the full potential of smart and reliable PNT information for the benefit of society as a whole.
For the ESA article on this topic, please click here.