Virtual assessment to enable safe automated driving
TNO, Siemens DISW, and Itility have started a TKI project to develop and demonstrate the use of virtual simulations for the safety assessment of Connected Automated Driving (CAD) systems. Case-by-case testing of individual functions of CAD systems is no longer sufficient to ensure operational safety on the road. This project enables Virtual Assessment of CAD systems using simulations of a large variety of tests based on TNO’s StreetWise technology.
Over the last years, TNO has developed a methodology, StreetWise, to build and maintain a real-world scenario database, suitable for testing and validation of CAD systems. Currently, StreetWise provides a solution on how to identify, parameterize and characterize scenarios from object-level in-vehicle sensor data provided from a fleet of vehicles driving on public roads.
In this TKI project, StreetWise+, TNO extends this method to identify and characterize scenarios using roadside-based camera systems. Moreover, TNO will integrate the effects of communication signals in the scenario definition, to accommodate assessment of communication-based automated systems, such as C-ACC (platooning), C-AEB (communication enhanced autonomous emergency braking), or Intelligent Speed Adaptation (ISA) functionality. Furthermore, TNO will extend the scenario models to allow multiple interacting actors to be part of the description of complex scenarios.
Itility and TNO partnered up in 2019 for StreetWise data analytics and data infrastructure (read more about our partnership in this blog). Here, sensor data from the StreetWise project, which is collected from cars driving in the real world, was processed in the Itility Data Factory. In this TKI project, Itility’s role will be to ensure efficient cloud-based data processing and develop a data stream quality check using their expertise in anomaly detection methods. Working closely together with TNO and Siemens DISW data scientists and software engineers, Itility will support the development of the statistical driver models and relevant APIs.
As the research proposed in this project has a good match with the HTSM Automotive Roadmap, the project received TKI funding. The project will run until December 2022.