SignalVoid

  Product Description 

Our solution addresses a critical challenge in drone navigation: ensuring reliable operation in GPS-denied environments, where GPS signals are either unavailable or severely compromised. Drones typically rely on GPS for navigation, but these signals are vulnerable to interference, spoofing, and jamming, which can jeopardize mission-critical operations. This is particularly concerning for rescue missions in dense forests, mountainous terrain, and polar regions, as well as in conflict zones where GPS spoofing and jamming are prevalent, such as in Ukraine or during airline signal jamming incidents in the Baltic region [1].

We offer a sensor-fused, AI-powered geolocation system with the goal of achieving meter-level accuracy. Currently, our solution delivers an average accuracy of 30 meters in GPS-denied environments. Cost-effective and easily integrate into existing systems with minimal modifications, our solution leverages satellite imagery and off-the-shelf, low-power hardware to ensure reliable navigation in challenging conditions. Designed to serve both as a GPS alternative and a redundant backup, our system enhances operational resilience and ensures continuity in mission-critical scenarios.

  Leveraging EU Space Technology

We leverage existing EU space data and Earth observation technologies to create a GPS-denied geolocation system. Our solution utilizes a camera equipped with a feature extraction pipeline, allowing the system to compare live aerial imagery with detailed Earth region data. This approach enables drones to predict their real-time location relative to the ground with high accuracy.

Specifically, we utilize Copernicus optical data from Sentinel-2 to cross-reference aerial imagery from Google Maps, with the latter serving as the reference for the autonomous aerial vehicle in this example.

For future developments, we plan to incorporate satellite data from the Satellite Portal [2] further to enhance the accuracy and reliability of our geolocation system.

  Addressing the EU Defense and Security Challenges

Our project offers a solution for unmanned drones tailored for defense and humanitarian operations (Challenge #2). The ability to operate in GPS-denied environments significantly reduces the risks associated with signal loss, whether due to jamming in conflict zones or other disruptions. By eliminating reliance on GPS, we enhance operational flexibility and situational resilience, providing strategic autonomy to EU defense forces, particularly in contested areas where traditional navigation systems may be compromised. This capability is vital for maintaining effective operations in humanitarian missions or military engagements in GPS-challenged regions.

  Future Development

The development will focus on implementing more advanced and intricate algorithms for feature extraction. This includes enhancing star tracking capabilities for nighttime navigation. Additionally, efforts will be directed towards optimizing the resolution of images captured by satellites, along with improving the efficiency of satellite image loading and processing.

  The Team

The team members all share a background at Forze, where they worked together on the software and electronics behind the world’s fastest hydrogen-electric race car.

  Benji Metz

MSc Microelectronic student at TU Delft focusing on hardware optimisations for AI. Researching deep learning for magnetically assisted 3D assembly. Worked as a manufacturing engineer at American aerospace company Rocket Lab and worked with ESA to launch REXUS31

  Dylan Durand

MSc Computer Science student at TU Delft following the Data Science technology track, which includes numerous specializations in AI. Currently researching a computer vision pipeline for autonomous marine robots.

  Timber Lock

MSc Quantum Information Science and Technology at TU Delft, which focuses on the development of quantum computing. Part-time working at Fox Crypto, developing hardware decryption devices for the defence industry.

  Luca Hagemans

MSc Computer and Embedded Systems Engineering, studying at the intersection between electrical engineering and computer science. Researching sustainability of IT systems at TNO.

  Martijn Crijnen

Currently working as investment advisor at ABN AMRO MeesPierson in the Investment Sales team. Co-Founder of StapStep BV, successfully launched, managed and sold that start-up, being mainly responsible for Finance and Sales.

  Business plan

  Our repository
https://github.com/BenjiMetz/CASSINI

[1] : https://www.airandspaceforces.com/russian-gps-jamming-nato-ukraine/. 
[2]: https://www.satellietdataportaal.nl/