Connected and autonomous cars represent the largest user community of high integrity high accuracy and multi-sensor positioning systems. A challenge - not yet fully addressed - in this domain is the validation and certification of multi-sensors positioning devices. This is a mandatory requirement for demonstrating the safety levels beyond current SAE Level 2 automated driving functions. This contribution - based on the P-CAR project funded by ESA on NAVISP Element 3 - aims at exploiting the process for achieving the safety levels of train control system, by promoting synergies and potential benefits for both automotive and rail application domains. Car manufacturers produce millions of vehicles and are investing in autonomy targeting higher safety. Meanwhile, the rail community has almost 20 years’ experience of managing train driving functions with a high level of safety to overcome dangerous driver’s behaviors as the European Rail Traffic Management System (ERTMS). P-CAR is realizing a competence center with laboratory facilities to evaluate the performance of car positioning systems by developing innovative simulators with Hardware-in-the-Loop (HiL) to create virtualized scenarios through a digital twin approach. That approach allows the evaluation of the UUT, reducing the time for field tests also constrained by limited reproducibility. The paper describes the use cases for relevant operational scenarios developed in cooperation with the automotive stake-holders. Yet it addresses challenges for the certification for which it is necessary to create realistic environmental conditions and proper fault injection strategy for evaluating the impact of events that have a rare occurrence and then are difficult to test. The reference multi-sensor architecture is the standard CEN/CENELEC EN16803 for using GNSS on road Intelligent Transport Systems (ITSs). The process of determining the safety requirements is defined according to the automotive standards ISO 26262 for functional safety and ISO 21448 for safety of the intended functionality. A peculiarity of the P-CAR laboratory is its virtualized cloud-based platform to involve other test-beds and laboratories creating a unique and skilled network.
File in questo prodotto:
Non ci sono file associati a questo prodotto.