Safe, durable trains require consistent maintenance. Technicians check the vehicles regularly; but during this process, tools might remain on the roof or maintenance flaps are left open. To avoid that, staff members currently have to do final visual inspections, so that the vehicles are safe and can go back in service. However, errors can occur due to time pressure and a tool is left around or an open flap is overlooked. VRVis and Zugkraft-kN address this issue in a joint project; together they design a digital roof scanner to support and optimize the inspections.
Automatic checks in live mode
For actual implementation, the application needs to be permanently installed and suitable for industrial environments. It must fit for various train models. Other relevant criteria are robustness, accuracy and qualitative suitability.
Digitalization of maintenance processes
The aim of this "Test before invest" project is to accelerate digitalization within the field of train maintenance. To achieve this, the research team will determine which hardware is suitable for actual use and test it under various conditions for different railway vehicles. Light and weather conditions, and the level of contamination play an important role. The project partners will evaluate the process they have already designed in the workshop of a customer of Zugkraft-kN. They will analyze whether the roof scanner should be equipped with 2D or 3D sensors. In addition, the team will examine the results of scans done of moving vehicles and whether models can be created based on the scans.
EDIH "AI5production" empowers companies to explore technologies
"AI5production" is an Austrian "European Digital Innovation Hub" (EDIH), embedded within a pan-European network. "AI5production" supports companies with technical expertise and gives the opportunities to experiment with technologies, for example via the "Test before Invest" funding scheme. The EDIH provides training and skills development, strengthening digital transformation efforts within companies.
Start: 01.2024
End: 06.2024
Zugkraft-kNGmbH
EC (Digital Europe Programme) and FFG (BMAW)
Digitization of construction sites: VRVis and clone:it automate rebar inspection and documentation.
The project REINFORCE researches how reinforcement learning and human-centered visualization methods can be used to solve complex control problems in an efficient, fast, and flexible way.
The HEROD project employs a combination of sensor technologies, data analytics, ontologies, and visual computing to improve predictability and quality assurance in 3D printing.
Construction sites are highly dynamic places, which makes complete and convenient documentation all the more important. As part of the "AI5production" program, VRVis and rmDATA are investigating whether mobile devices with LiDAR and high-resolution cameras are suitable for automatically recording construction processes in 3D.