A new research and development project led by Geospatial Ventures Limited (GVL) has introduced groundbreaking artificial intelligence (AI)-driven solutions for high-precision positioning, navigation, and timing (PNT) in sensitive environments. These innovations aim to enhance accuracy in locations that demand exceptional monitoring, such as brownfield sites, hazardous terrains like peat bogs, and aging or newly constructed infrastructure, where ground shifts and other disturbances can have significant implications.
At a recent presentation organized by the European Space Agency (ESA), GVL’s Managing Director, Paul Bhatia, highlighted the limitations of existing technologies. Traditional PNT solutions, such as conventional PNT systems and real-time kinematic (RTK) base stations, are often costly and ill-suited for diverse applications. Bhatia explained that AIPLAN, the new system developed by GVL, integrates a customized, commercial off-the-shelf (COTS) GNSS receiver with InSAR (interferometric synthetic aperture radar) technology, enhanced by AI and machine learning (ML) to deliver superior performance.
InSAR, a radar-based technique that detects ground deformation by analyzing Earth’s surface via satellite images, forms a critical component of the AIPLAN system. The AIPLAN setup consists of geodetic receivers and specially designed rovers that collect GNSS data, while InSAR data is gathered through corner reflectors and other natural reflectors. All data collected is uploaded to the AIPLAN platform, where AI/ML algorithms are used to smooth errors and train the system through pattern recognition and anomaly detection.
Targeted Precision with Advanced Features
The AIPLAN project aimed to achieve positioning accuracy within 1-5mm horizontally and 2-5mm vertically. To maintain cost-efficiency, low-cost GNSS modules were integrated into the system, ensuring affordable hardware without compromising performance. The system is also ruggedized and waterproof, making it suitable for deployment in challenging environments. Designed for real-time monitoring, AIPLAN provides daily position updates, with easy-to-use software accessible via mobile phones or laptops.
The project’s testing phases took place in several locations across the UK, including the Spen pig farm at the University of Leeds, Snake Pass on the A57 in the Peak District, and the Very Light Rail National Innovation Centre in Dudley, West Midlands. These trials involved calibrating and evaluating GNSS base stations, rovers, and dual InSAR corner reflectors.
Positive Results and Future Applications
The results of the tests were highly encouraging, with AIPLAN systems achieving positioning accuracy of 3-4mm in both horizontal and vertical measurements. Additionally, the newly developed visualization suite provided invaluable time-series data for continuous monitoring of key points. According to Bhatia, the AIPLAN technology is versatile, making it suitable not only for infrastructure monitoring but also for tracking geohazards such as landslides and earthquakes.
Funded by ESA’s NAVISP program, which supports innovation and competitiveness in Europe’s PNT industry, the AIPLAN project marks a significant step forward in the use of AI and machine learning for precision land planning and monitoring in critical areas.
Related topics:
Tech Mahindra and AWS Launch AI-Powered Network Operations Platform
Canadian News Outlets Sue OpenAI for Copyright Infringement
Elon Musk Files Legal Motion to Block OpenAI’s Shift to a Full For-Profit Model