AI technology that “understands” technical concepts is being used to help press users search for the parts they need.

With millions of parts to choose from, manroland Goss is leveraging the Partium mobile app – as well as a PC version – to enable customers identify a part without any descriptions of it, by using a combination of image, semantic text and a parts list search.

With some complex systems, there can be up to 40,000 parts installed in up to 500 assemblies. The company says this can result in up to 10,000 consumable parts that have to be replaced regularly.

The job of researching spare parts is simplified using the Partium application for mobiles and for PC, speeding up reordering of ‘spare and wear’ parts with patented machine learning. Additionally, a ‘call-an-expert’ function enables direct contact with the maker to resolve queries.

Partium’s image search makes it easier to identify parts, recognising them in disassembled and assembled states, and subsequently enables precise search results through filter options such as designation, dimensions, weight and parts list information.

Going beyond simple image matching, the AI technology ‘understands’ the technical concept of the part being searched for and its intended use, and uses data for millions of parts which have been analysed in advance.

It also enables a smart search by text input, using just a description of the spare part being searched for. Numerous filter options according to ID, brand name and dimensions also allow a quick search.

Customers can also navigate to the spare part they are looking for by focusing on those for a specific press, or even submit photos.

A demonstration is available on the company’s website.