Multi-sensor fusion techniques are the next challenge for resilient and high accuracy – integrity positioning solutions for the ERTMS train control and the Connected car applications. Radiolabs – leveraging on technical synergies among these sectors – and its University-Industry research organization will present the latest achievements at the prestigious ITM 2022 conference bringing together international experts.
The International Technical Meeting – ITM 2022, organized by the Institute of Navigation (ION) with a technical program related to positioning, navigation, and timing has selected and peer-reviewed three papers to be presented in Long Beach January 25-27.
The first paper – On the Design of High Accuracy Rail Digital Maps based on Sensor Fusion – is relevant to the adoption of multi-sensor train positioning solutions for the ERTMS applications. Images, depth maps, and point clouds are used to complement an IMU/GNSS localization module with a sensor fusion-based technique for the digital map design. This research is part of a collaboration between Radiolabs with the University of Roma Tre, University of Padova, and RFI who is pioneering the validation and certification of the first GNSS-based ERTMS system in Italy on the Novara-Rho Pilot line.
The second paper – On the Validation of Multi-sensor High Integrity Positioning Solutions for the Connected Car: the P-CAR Infrastructure – deals with the realization of P-CAR laboratory, a facility conceived by Radiolabs to support the certification of multi-sensor positioning devices for tier-2 solution providers and car manufacturers beyond current SAE Level 2 driving functions. The vision is an open and independent laboratory based on a virtualized platform relying on the Edge and Cloud paradigms to create a geo-distributed network of laboratories. Virtualization and zero-on field tests are the main feature of P-CAR to create digital-twins of cars interacting with the operational scenarios to assess the safety requirements. The research is being conducted in strict collaboration with the University of L’Aquila.
The third paper – A Semantic Segmentation-based Approach for Train Positioning – is presenting a novel approach to overcome the vulnerability of GNSS-based high integrity train positioning to signal degradations and interferences. A fusion of data from GNSS and visual sensors is proposed as a means to provide a trusted, reliable and, continuous train position in challenging railway environments. Different sensors are combined to increase the accuracy and to enhance the robustness of the overall system while the redundancy gained through the multi-sensor fusion approach will be crucial to overseeing the integrity requirement for the safety-critical rail applications. This research has been pursued by a team of Radiolabs, University Roma Tre, University of Padova.
These results are achieved thanks also to the support of the space agencies: ASI, ESA, EUSPA.
