Work

Towards Virtual Testing of Perception in the Railway Domain

Year: 2025

Type: conference-paper

Abstract: To achieve higher automation levels in the railway domain (GoA-3/4), especially focusing on branch lines, an automated system must take over responsibilities which are currently allocated to the train driver. While existing automation technologies ensure operational safety at the macroscopic and technical levels, addressing perception at the operational level in open context remains a challenge. This paper describes key aspects of future railway perception components, and identifies challenges that they bring for verification and validation. It is shown how existing technologies may be combined into a holistic framework that finally enables scenario-driven simulative testing of perception components.

Source: IEEE International Automated Vehicle Validation Conference (IAVVC)

DOI: 10.1109/iavvc61942.2025.11219512

Authors: Michael Wild, Jan Steffen Becker, Anda Buinoschi-Tirpescu, Eike Möhlmann

Institution: Deutsches Zentrum für Luft- und Raumfahrt e. V. (DLR)

Language: English

Topic: Transportation Systems and Safety

Subfield: Computer Graphics and Computer-Aided Design

Field: Computer Science

Domain: Physical Sciences

SDG: Industry, innovation and infrastructure

Open Access status: green

Funders: European Commission, HORIZON EUROPE Framework Programme