TRA, the Transport Research Arena, is the foremost European transport event that covers all transport modes and all aspects of mobility. It is the largest European research and technology conference on transport and mobility with 4000+attendees and 150+ exhibitors at the 2024 event in Dublin.
Alessia Vennarini from Radiolabs gave a presentation on the latest achievements of the RAILGAP project funded by EUSPA. RAILGAP team is targeting an advanced solution to developing methodologies and tools for building high accuracy and precision Ground Truth and Trackside Digital Map for the railway environment.
Ground Truth provides a set of reference data with a bounded accuracy to be used for determining train position, speed and acceleration without a priori knowledge of the database of the railways lines. While the Trackside Digital Map generates and maintains a digital twin of the railway network with updated signalling Informations and track layout.
The starting point is to assessing high fidelity ground reference points along the railways by exploiting data fusion algorithms between EGNSS signals and IMU, LIDAR and Camera data gathered by a set of sensors installed on commercial trains in routine services. The recorded data are then processed by Artificial Intelligence (AI) techniques.
Radiolabs is responsible of the trackside digital map toolset methodology definition, development and verification to detect and locate relevant objects (e.g. signalling track elements) within railway scenes using deep learning algorithms and to automatically create the Trackside Digital Map. The track elements detection and processing is based on algorithms relying on data-driven automatic learning strategies applied to the collected data. In addition, the toolset is able to perform the continuous monitoring and control of the Trackside Digital Map to detect critical deviations with respect to the version assumed to be the reference Map. RAILGAP Project
