AI Navigation System Offers GPS Free Accuracy

Driving without GPS might seem like a step backward, but a groundbreaking AI system from the University of Surrey could make it the future of navigation. Researchers have developed an innovative tool that can pinpoint a device's location with remarkable accuracy, even in dense urban areas where traditional satellite signals often fail.

This new system, called Pose-Enhanced Geo-Localisation (PEnG), demonstrated its power in tests by reducing location errors from an average of 734 meters down to just 22 meters. Instead of communicating with satellites, PEnG intelligently combines satellite imagery with street-level photos to determine both a device's location and its orientation. The process first establishes a general position at street level and then refines it using relative pose estimation to calculate the camera’s direction. Remarkably, this advanced capability works with a simple monocular camera, the type commonly found in modern vehicles, making it a practical and accessible technology for widespread use.

A Smarter Way Forward

The primary motivation behind PEnG is to address the unreliability of GPS in challenging environments like tunnels, between skyscrapers in cities like New York, or in any region with poor connectivity. The system was specifically designed to fill these crucial gaps.

"Many navigation systems depend on GPS, but coverage isn’t always guaranteed," explained Tav Shore, a postgraduate research student in AI and computer vision. "Our goal was to develop a solution that works reliably using only visual information. By combining satellite and ground-level imagery, PEnG achieves a level of accuracy previously thought unachievable without GPS – and could help unlock new possibilities for autonomous vehicles and smart navigation tools."

Dr. Simon Hadfield, associate professor in robot vision and autonomous systems, added that the system’s design is a major advantage. "One of the most exciting aspects of this system is how it turns a simple monocular camera into a powerful navigation tool," he said. "PEnG is designed to operate without GPS, making it ideal for fast-moving, unpredictable scenarios. That kind of flexibility is exactly what’s needed for the next generation of autonomous vehicles and robotics operating in challenging environments."

Beyond GPS Dependence

Our global reliance on GPS means that any outage or signal interference can have significant consequences for industries ranging from logistics and aviation to defense. GPS signals are vulnerable to disruption from tall buildings, tunnels, jamming devices, and even atmospheric conditions. By offering a robust, GPS-free alternative, PEnG could significantly improve the resilience and reliability of navigation systems worldwide.

Professor Adrian Hilton, director of the Surrey Institute for People-Centred AI, emphasized the project's practical impact. "Our team’s work demonstrates the people-centred approach to AI we champion here at Surrey, developing a system that addresses the challenges behind navigation technology, something we’ve all come to rely on," he noted. "The ability to accurately pinpoint a location without GPS lays the foundation for smarter, more resilient autonomous systems that can operate in even the most remote environments.”

The project, supported by a University of Surrey PhD Foundership Award, is now moving toward the creation of a working prototype for real-world testing. To foster collaboration and accelerate progress, the team has released their research as open source, allowing developers globally to build upon their work.

The complete findings have been published in the scientific journal IEEE Robotics and Automation Letters.