IT Platforms and Cloud
A Secure and Future-Proof Data Processing Infrastructure
Digital railway technologies demand a new and safe data processing infrastructure for use on trains and track. Oriented on the latest trends in cloud computing, the technology still meets stringent railroad standards. Digitale Schiene Deutschland relies on technologies such as artificial intelligence, high-precision positioning systems and environmental perception using sensor technology to control train traffic even more efficiently in future and guarantee journeys at optimal intervals. This requires large amounts of data to be processed in real time. For example, trains must be able to determine and transmit their location continuously, reliably and with absolute accuracy during their journey. The train must be able to pass on information about speed, acceleration and train completeness, to the control and safety technology and traffic management system that schedules nationwide railroad operations. This requires a powerful, scalable and future-proof data processing infrastructure.
The Advantages at a Glance
An essential foundation for an efficient, scalable and future-proof data processing infrastructure is to firstly clearly separate applications (e.g. environmental perception and incident management) from the underlying data processing infrastructure or platform. Such separation – and the introduction of a standardized interfaces – allows the very different life cycles of applications and platforms to be taken into account and both sides to be migrated independently of each other. Furthermore, in this way the railroad sector as a whole can draw on a larger market of application and platform manufacturers and achieve better cost efficiency through the use of cross-industry standard software and hardware.
Against this background, the concept for a “Safe Computing Platform” was developed in 2021/2022 by the European “RCA” and “OCORA” initiatives and in collaboration with partners such as Siemens Mobility, Thales and SYSGO. This allows applications and platforms to be sourced from different manufacturers, while still meeting the high safety requirements (up to SIL4) demanded by critical applications.
The digital railroad system of the future not only places stringent demands on the data processing infrastructure, but also has very different and individual needs. Applications based on artificial intelligence require data processing infrastructures with hardware acceleration, and applications that are particularly latency-critical should be placed close to the action, i.e. in so-called ‘edge clouds’. Applications running aboard trains, on the other hand, have completely different requirements due to factors such as operational constraints aboard the train, higher cost pressure, less space, etc. Taking these diverse needs into account, Digitale Schiene Deutschland is developing a hierarchical data processing infrastructure consisting of private central, regional and local cloud-like data centers, combined with a powerful new onboard computing platform.
The Technologies in Action
A Digital Rail Testbed in the Ore Mountains shows how integrating and validating digital railroad system applications based on FRMCS/5G and cloud technology works in practice. An important component of the digital test field was the construction of a mobile communications base infrastructure on a test track around ten kilometers in length. This forms the basis for testing FRMCS/5G and cloud equipment from various manufacturers.
Alongside data transmission and processing, digitization also demands the systematic generation and provision of data in order, for instance, to be able to adequately “train” artificial intelligence. This will play a decisive role in many future use cases. This requires a so-called “Data Factory” that systematically generates and provides large quantities of the necessary data for “training” AI software, among others. To this end, an infrastructure consisting of state-of-the-art IT platforms with very large storage units and high-performance computers, together with a cloud environment, has been under construction since 2022. The Data Factory project highlights how IT platforms and cloud solutions are used in practice.