Research projects

In public transport and freight transport, many different trains are on the move on Austria's rails. These must be regularly maintained and damaged parts quickly replaced. Particularly in the event of a break-down, it is important to have the necessary spare parts on hand quickly in order to keep downtimes as short as possible. However, it is a time-consuming and costly task to actually have all spare parts of all trains in stock at all times. Factors such as storage space, transport emissions, sustainability and availability also play a major role. Spare parts for older machines may no longer be produced at all. All of this poses major challenges for transport companies - especially in times, when they look out for climate-friendly options.

A research team led by Fraunhofer Austria is now working with ÖBB Train Tech, Wiener Linien, Wien Energie GmbH, VRVis and Fraunhofer IAPT on a solution to produce train spare parts using 3D printing. To this end, thousands of technical drawings, some of which were made by hand, are being analyzed and evaluated to determine whether they are suitable for additive manufacturing. Specifically, the researchers are analyzing the technical drawings using image processing and pattern recognition as well as text recognition and text mining, and then linking them to other data sources. The research goal is to read out the contents of the drawings in a standardized and automated way. A life cycle assessment of the spare parts will then provide information on whether 3D printing makes economic sense in the respective case. Subsequently, a technology assessment will clarify whether the state of the art of 3D printing can meet the necessary requirements for the components, for example in terms of strength, stability or fire protection. Finally, the potentials at ÖBB and Wiener Linien are to be bundled in the sense of a sustainable sharing economy in order to use a 3D printer together as effectively as possible.

In the AM4Rail project, VRVis is researching and developing the pre-processing pipeline which, by means of image processing and semantic analysis of 2D images of 3D objects, allows geometric properties to be derived as automatically as possible from the large quantity of partly historical, hand-made digitized technical drawings, which contribute to the technological evaluation of spare parts with regard to 3D printing (additive manufacturing).