How Visual AI Can Help Industries Overcome the Talent Gap

In 2022, researcher Mimi von Blomberg from Tallinn University of Technology conducted a study at nyris showing that visual search is one of the most powerful AI technologies within this shift. Technicians are able to find what they’re looking for using photos and AI instead of relying on catalogs, customer service hotlines, or that senior expert employee.

The study’s goal was to determine whether or not visual search engines would allow people to identify spare parts, components and products faster than text-based search methods. Additionally, von Blomberg wanted to investigate whether the user experience of using visual search was preferable to experts in the industry.

Back in May 2022, von Blomberg gathered a group of 55 individuals to take part in her study. The participants were split into two groups: experts and non-experts. The 23 people in the expert group had experience as service technicians or spare parts managers, and had spent at least 8 hours weekly identifying spare parts. The other 32 were “laypeople”—mechanical engineering students or marketing and sales employees who had little direct experience with spare parts.

The participants were tasked with identifying 16 spare parts out of 35,000 from the Ganter Norm Parts Catalog, either using keyword-based search or visual search methods. Von Blomberg measured search time and search success, meaning whether or not the participants were able to correctly identify a spare part.

The data von Blomberg collected showed that visual search had a clear advantage over text-based search. On average, the participants found the right spare part 3.3 times faster with visual search, and they were 51% more successful at finding the part. In addition, 94.6% of the participants reported that they would like to apply visual search in their work environment, and 89.1% felt that visual search was better than text search.

These findings were most significant for the non-expert group, who were significantly better at finding spare parts by both metrics using visual search. This has big implications for visual search—you will no longer need an expert to identify and locate spare parts. As today’s economy faces challenges replenishing its expert workforce, this is a great solution that industrial manufacturers can implement immediately.

The German Association of Mechanical and Plant Engineering, or the VDMA, is becoming increasingly concerned with a lack of skilled personnel. Technical industries are experiencing unprecedented consumer demand and thus high industrial growth—in Europe, the mechanical engineering sector is expected to grow at a rate of 3.8% annually for the next 10 years (1). Yet, as professionals from the “baby boomer” generation retire and the modern world experiences a cultural shift away from trade skills, younger generations aren’t producing enough workers with the right education to keep up with the booming industry.

This skill shortage is a serious issue that has industry leaders holding their breath. In a June 2022 survey, the VDMA found that 60% of mechanical engineering companies consider the number of skilled workers to be a big risk going into the future—even bigger than inflation and geopolitics (2). Without experts on board to maintain lean technical operations, manufacturing companies risk falling behind in the context of the global market.

Two thirds of a customer’s total cost of ownership comes through after-sales—successful OEMs (original equipment manufacturers) make more money providing services and spare parts than they do from selling machines themselves. However, it’s becoming increasingly easier to get access to whatever spare parts you need through other, much cheaper sources. Companies will struggle as people turn away from their after-sales in favor of more affordable options from the online retail giants that dominate the worldwide market.

Where manufacturing companies have the upper hand, however, is the service aspect—they can provide thorough and efficient support that the customer would not find from ordering parts cheaply online. They can stay globally competitive if they optimize their after-sales services to accommodate their large customer demand.

‍

As von Blomberg’s research demonstrated, visual search is one of the most feasible and market-ready solutions to this problem. It enables anybody to identify spare parts quickly and accurately, even those without expertise. The way after-sales services function has always changed as technology evolved.

Decades ago, technicians referred to catalogs when searching for spare parts. Down the line, they converted to service hotlines, emails, and clunky asset management systems. Specialists would spend fortunes flying out to wherever they were needed to service industries with their expertise. The coronavirus pandemic prompted a surge in remote assistance, which made travel, in many cases, obsolete. Today, technology has advanced far enough to put all of that behind us; the current shift is moving us into the realm of AI, with visual search being one of the first products to enter the market.

Historically, when shifts like this occur, the market players that don’t react early end up crumbling. Take Kodak, for example—after ignoring the shift to digital photography, they have all but collapsed. Even after inventing digital photography, knowing it would be the next sensation, and having a decade to prepare, executives in the late 1980s decided to stay close to their roots in film (3). Kodak’s film-based products started to flop while other companies like Canon, Nikon, and eventually Apple, who invested in digital photography, surged forward and took over their space.

Given the lack of skilled personnel in manufacturing industries, the prominence of visual search is inevitable. It will be up to individual companies to decide whether they will adopt visual search early on or a decade down the line—technology will advance with or without them.

It’s clear that visual search provides massive benefits over keyword-based search, and that everyone will benefit from it, whether they have years of experience with spare parts or no experience at all. After all, even the experts who participated in Blomberg’s study benefited from using visual search—they experienced a significantly higher success rate using visual search over text-based search, and also an increase in speed. What’s even more impressive is that almost 95% of the overall participants shared that they would like to implement visual search at work.

Not only do employees want visual search, but their companies will need it in order to compete in the global market. As the next wave of technology arrives, visual search will be one of the most essential tools to overcome the challenges of today’s industrial world.

‍

Sources:

1. https://single-market-economy.ec.europa.eu/sectors/mechanical-engineering_en
2. VDMA
3. https://www.forbes.com/sites/chunkamui/2012/01/18/how-kodak-failed/?sh=4a57e8de6f27

‍