As future cars will likely become a supercomputer on wheels powered by AI massive data-collection and traditional roads will evolve to smart infrastructures with mobility services provided to the drivers, processing tools are the new engine for trusted vehicle’s positioning, that is undoubtedly the gatekeeper to adoption.
A remarkable progress in the fields of Automatic Driving Systems (ADS), together with mounting interest for machine learning (ML) and artificial intelligence (AI) technologies, have increased the importance for massive data-driven applications to assess a trusted positioning of the car for the partial or complete control of safety-critical operations. In this context a great importance regards the ability of the ADS positioning function of a car to safely execute its task, which is challenged on how the integrity of the position data is guaranteed whatever is the operational environment, the weather and road conditions. A trusted position of vehicles is also the pre-requisite for the road infrastructure manager to provide smart road services that are a unique cross-sector ecosystem bringing together vehicle manufacturers, service providers, public authorities, and associations to transform vehicle-generated data into life-saving insights. By sharing real-time safety-related traffic information European roads will become safer contributing to the EU’s Vision Zero ambition.
In the realm of modern automotive testing and validation, the integration of sophisticated data processing and simulation systems is paramount to ensure the safety of ADS, which must rely on the knowledge of the position of the car and the surrounding environment. Positioning and Sensing are both dependent on the car sensor system and the processing process which in turn depends on the data gathered on the field.
STEV – Satellite Testbed for Egnss based vehicle localization Validation – supports this integration, allowing advanced positioning modules to be tested under controlled yet highly realistic conditions. The system conceived by Radiolabs is a sophisticated framework designed for the rigorous testing and validation of algorithms to fusing data from different sensors. The system comprises three main components: a multi-sensor On-Board Unit, a Central Computing Unit and a Replay Unit. Together, these components form an interconnected architecture that supports the post-test analysis. The system’s ability to simulate real-world conditions and to perform thorough verifications under controlled environments significantly enhances the reliability and performance of the algorithms being tested. The goal is the design and validation of the positioning algorithms for meeting the SAE level 3 of automation (the first to allow autonomous capability with a driver ready to take control in case of ADS disconnection) up to level 4 and 5 where the presence of the driver is not necessary.
The multi-sensor On-Board Unit is responsible for gathering data from various sensors within the vehicle, including GNSS receivers, Inertial Measurement Units (IMUs), LiDAR and Camera sensors that contribute to the accurate localization and monitoring of vehicle dynamics. These data are then transmitted to the Central Computing Unit, to undergo extensive processing and analyses. The Central Computing Unit acts as the brain of the STEV system, employing advanced algorithms and computational techniques for the generation of test patterns to be used to evaluate the performance of the modules under test. Finally, the Replay Unit allows for the reproduction and detailed examination of specific test scenarios, enabling engineers to scrutinize the system’s responses and identify any potential areas for improvement.
The STEV project has successfully passed the Critical Design Review – CDR, meaning that the selected architecture is ready for the implementation phase. The CDR data package is a comprehensive blueprint, guiding the seamless integration and operation of various subsystems, ensuring they work harmoniously to achieve the desired performance and safety standards. The roadmap has been frozen for planning the system completion with rigorous testing and reliable validation processes, forming the backbone of the STEV system.
The project has been commissioned by ESA – under the GSTP programme funded by ASI -and is coordinated by Radiolabs, with Intecs as partner. ANAS – the Italian Road Infrastructure manager with 32,000 Km of roads – is contributing to STEV with their use cases and by supporting the test campaign with their vehicles collecting data on selected roads that are representative of their Smart Road Initiative. Honda Italia too is contributing to the project to improve the safety of motorcycles and a motorcycle, will participate to the test campaign. Both vehicles provided by Anas and Honda Italia will be equipped with the multi-sensor on board unit and the data collected are processed by the central computer unit.
STEV is a pillar of the Radiolabs strategic plan for developing and validating the enabling technologies for the connected and autonomous vehicles, contributing so far to an eco-sustainable, safer and more efficient transportation system in line with the priority of Europe to reduce road fatalities by 30% in 2030 and zero fatalities in 2050.
