
Image recognition partnership creates part ID company
The newly formed company, Partium, offers a solution for the fast identification of parts within industrial and retail environments.
“Our companies each developed compelling and complementary offerings for part recognition,” says Humai CEO Philipp Descovich, who will become CEO of Partium. “Humai is the leader in spare part recognition for industrial and manufacturing industries. Slyce is the leading provider of visual search technology for retail clients like The Home Depot and Leroy Merlin, as well as a new in-store solution, Part Finder kiosk. Catchoom has long-standing experience in large-scale image recognition as well as AI-based solutions for enriching product metadata.”
All of these capabilities will be provided by Partium, say the companies, offering an end-to-end solution to the challenges in the identification and ordering of spare parts, and enabling their business to power a larger addressable market. The new company will build on Humai’s part-recognition solutions that combine computer vision with “Delta,” a proprietary spare part AI, to achieve a 95% or better exact-match recognition rate of industrial parts.
Humai’s customers, including Bosch, Daimler, and Deutsche Bahn, leverage this software in mobile apps to identify installed parts as well as loose parts in a warehouse.
“With our part search, we can dramatically decrease the time it takes a technician or consumer to identify a part, and also avoid misidentification and mix-ups,” says Descovich. “For machine builders, Partium is a new and exciting channel to grow their aftersales business through promotional offers, parts catalogs, and user data analytics that help to predict demand.”
“The ROI is immediate and measurable,” says Descovich, “not just for our current industrial and manufacturing clients, but also for the automotive and home improvement retailers that Slyce has a history of working with. With Catchoom and Slyce, we have formed a team that can achieve anything.”
The Slyce company, a leading provider of visual search for more than 60 retailers in North America, developed its expertise in part recognition with The Home Depot, which launched Slyce’s camera search in 2015. In addition to rapid consumer adoption of the feature, The Home Depot has highlighted Slyce’s part recognition in several national TV ads. Tractor Supply Company, NAPA Auto Parts, Leroy Merlin, Hayward Pools and AES have all partnered with Slyce around part recognition in recent months, says the company.
“Identifying spare parts – whether it’s a hex bolt or a link-arm assembly – is hard, especially if you don’t have that domain expertise,” sats Ted Mann, CEO of Slyce, who will become president of Partium. “What we realized at Slyce was that our technology was exceptionally good at getting to an exact part match.”
“At first,” says Mann, “we did this only with a mobile experience in home improvement and automotive retailer applications. But increasingly, we’re using in-store solutions, like our Part Finder Kiosk, which can truly find a needle in a haystack, identifying any fastener you put inside and blinking a light to show you exactly where to find it in the aisle of a store.”
Catchoom, an award-winning image and object recognition provider, recently developed a robust, fine-grain image classification engine and product information management service, which enables clients to automatically enrich product data. Metadata management is one of the largest challenges around part identification.
“Slyce, Catchoom and Humai were each solving different pieces of the part search problem,” says Catchoom CEO and Co-Founder David Marimon, who will become the Chief Product Officer of Partium. “By integrating the best capabilities from each, we can have the most complete and powerful solution. Together, we combine the tools to visually identify parts, collect training data, enrich metadata beyond customer’s own catalog information, and build embedded solutions. Whether you’re trying to identify a part in a warehouse or a store or a machine, we’ll have the end-to-end solution.”
According to Descovish, Partium’s technology will be designed to be modular and platform agnostic.
“In other words, you can recognize parts using our Partium app, through the Partium SDK embedded in your own ERP app, a Partium kiosk that lives in the warehouse, or however you chose.” he says. “Our goal is to make your spare part catalog and access to expert support truly available at the shop floor.”
All part-recognition solutions will be developed on the Partium platform, the company’s flagship product. The company will continue to support customers outside of part recognition through its other products, including the Slyce SDK for visual search in fashion, grocery and home decor; CraftAR SaaS for Image Recognition and AR; and MagicLens for AR renderings in sales materials, trade shows and presentations.
The Partium business will be headquartered in Vienna, with offices in Philadelphia, Barcelona, and New Waterford, Nova Scotia. Slyce’s lead investor, Anzu Partners, and Humai’s lead investor, Berndorf AG, will be the two largest shareholders in the merged company.
In addition to CEO Descovich, President Mann, and CPO Marimon, the leadership of the company is rounded out by Chief Technology Officer Adrian Ion, Chief Science Officer Tomasz Adamek (former CTO and Co-founder at Catchoom), and Chief Financial Officer Christoph Lehner. The combined team will have 90 employees, including what it says is the largest group of seasoned engineers and PhDs in the world that is solely focused on solving the spare part search problem.
Related articles:
AI-based brand protection platform takes aim at counterfeit listings
Optical AI product ID startup aims to boost trust in supply chains
AI-driven on-demand product authentication startup raises funds
Food recognition AI turns photos into recipes
AR smart glasses to merge with AI, computer vision to ID images
