European Common Data Management Platform Definition for Railway AI Function Development

Mikel Labayen,Daniel Ochoa de Eribe,Ander Aramburu,Marcos Nieto,Naiara Aginako
DOI: https://doi.org/10.55121/tdr.v2i1.143
2024-03-12
Abstract:Digitalisation and automation of operations in the railway industry include the use of Automatic Train Operation systems that provide automated functions to reach different levels of automation, known as the Grade of Automation (GoA) levels. These levels go up to GoA4 in which the train is automatically controlled without any staff on board. Artificial intelligence has emerged as technology that can substitute humans in certain driving tasks, in GoA3 (driverless) and GoA4 (unattended) modes. AI capabilities include perception, decision-making, precise positioning, or optimization of communications. The success of AI models depends on the quality and diversity of the data used for training, along with the set-up of a data life-cycle framework that covers creation, training, testing, deployment and monitorisation. The management of training datasets implies both expensive and time-consuming data gathering, labelling, curation and formatting efforts, potentially hindering the development of reliable AI systems. This paper presents a Common Data Management Platform developed by a consortium of European railway stakeholders, devised to efficiently manage data for AI training, and which is demonstrated in two different Proofs of Concept.
What problem does this paper attempt to address?