Georeferenced acceptance testing of LST elements: Pilot project in Mannheim highlights potential for the digital future
Today, acceptance testing of control and safety technologies (LST) often involves a great deal of manual work, track closures and work on the tracks. A pilot project at the Mannheim train formation facility demonstrates how this process can be carried out in future with significantly less effort: Using high-precision multi-sensor detection, AI-supported data analysis and the DRVito software, LST components are recorded with georeferenced data and cross-checked against the planning data. The aim is to make inspections faster and more efficient and, in the long term, to enable a large proportion of the acceptance process to be carried out from a desk.
How can acceptance processes in control and safety technology be made faster, safer and more efficient? This question is becoming increasingly important in the context of the digitalisation of rail transport. This is because infrastructure projects are subject to significant time pressure, whilst testing and acceptance processes – which are often still carried out manually – are resource-intensive and involve work on the tracks. A pilot project at the Mannheim train formation facility demonstrates how a new approach can help in this regard: the georeferenced acceptance testing of LST elements.
The aim of the pilot project was to further digitise what had previously been a labour-intensive testing process and, in the long term, to develop it so that a large part of the acceptance testing can be carried out ‘from a desk’. The focus is on the digital comparison between the actual infrastructure and the corresponding design data. To this end, LST elements such as signals, axle counters and balises are recorded with high precision, documented with georeferencing and subsequently evaluated digitally.
The technical basis is provided by multi-sensor data collection, in which modern measurement systems capture the infrastructure and make it available as a digital database. The information collected is used, amongst other things, to generate point clouds that precisely map the actual situation on site. AI-supported methods help to automatically recognise and classify relevant objects. The coordinates of the installed LST elements can be determined immediately after installation using a GNSS surveying instrument. This uses satellite signals to determine precise positions on the Earth’s surface. The data is then transferred to the DRVito software. The DRVito software plays a central role in the evaluation: it supports the comparison of planned versus actual outcomes and enables virtual site inspections of the facility. This allows verification of whether LST elements have been correctly positioned and implemented in accordance with the plans.
The project expands the so-called ‘Consistent Digital Data Management in CCS Planning’, or D3iP for short. The aim is to link planning, construction and operational data more consistently. Data discontinuities are reduced, data quality improves and verification processes can be better standardised. The pilot project thus lays an important foundation for more efficient acceptance procedures and an accelerated roll-out of modern control and safety technology.
The results from the pilot project in Mannheim show that georeferenced acceptance testing is feasible in practice and offers great potential. It can reduce track closure periods, minimise testing efforts and make infrastructure projects more transparent. At the same time, it enhances safety for staff, as fewer tasks need to be carried out directly in the track area. In the long term, this will lead to a new vision: planning, construction, acceptance testing and operation will be digitally integrated. Acceptance testing from a desk could thus become a key component of the digital infrastructure inspection of the future.
Further details on the topic of georeferenced acceptance testing of LST elements can be found in a recent technical article in the magazine "Deine Bahn" (available in German only).
