3D scanning and reverse engineering algorithms in the railway sector

Dette prosjektet har tatt frem utstyr og metode for å 3d skanne deler til tog og andre bransjer. Man har kommet et stykke på vei til å lage en direkte 3d modell av denne typen skann via formgjennkjenning. Det betyr at dataprogrammet forsøker å "skjønne" hvordan delen ser ut og dermed lage en tegningsmodell man kan endre på. Behovet for denne teknologien i referanse til jernbanen er stort, og man ser flere muligheter for å gjennvinne deler og holde eldre tog i trafikk lengre. Prosjektet har stoppet opp da det viser seg meget vanskelig å få godkjent denne typen deler i jernbanen da resten av systemet ikke er modent nok. Men teknologien og metodene kan gjenbrukes og videreutvikles for både jernbane og i andre bransjer.

-

The transport sector, including rail, depends on rapid access to spare parts so that transport services can be delivered with the required regularity. Original equipment manufacturers (OEMs) guarantee supply of spare parts only for a limited number of years and have long lead times. Lack of competition implies a high cost for spare parts even when they are accessible. Technical information, including detailed parts documentation, is often withhold. Ageing fleets of trains where parts obsolescence is commonplace often make it difficult to keep vehicles in service. The project goal is to establish a reverse engineering workflow from physical parts to well behaved CAD models facilitating rapid manufacturing of replacement parts as well as a library of digital replica and tools for training personnel and efficient maintenance of components. Two prototype products are envisioned: A precise, automized and user-friendly photogrammetry scanner, and a tool to enhance point cloud data to CAD models compatible to the ISO 10303 format STEP and fit for subsequent use. The latter will be mainly automatic but require user input for quality control. Algorithms for reverse engineering have a long history, but the tools used for industrial applications are far from automatic. Selecting and extending new and existing algorithms and workflows to rise the user friendliness and degree of automation involve challenging research. To integrate the new tools in the maintenance work, the right parts to be reverse engineered must be selected. The results will be consolidated in a roadmap starting from part selection to reverse engineering of parts.