AI-industrie.png

Smart Manufacturing & Materials-iStock-1135159658-BEWERKT.jpg

<p><strong>“Artificial intelligence (AI) is often used as a buzzword in the industry, but I don’t agree with that. You can do amazing things with it, but it’s not a magic bullet,” says Arno Dumoré, owner of Duflex Mechatronics, the company that has been building machines and solutions for various sectors for 25 years. “We’ve been applying AI for the past five years. Our goal? To make those applications just as smart as those in an iPhone, if possible. Because, yes, the iPhone is ahead of the industry.”</strong></p><p><img src="https://cms.aihub-oost.nl/uploads/MPTW_2xea_RS_1_V_Id_Wy_D_Wvk_A_Uzz_Mc9_L_Shr_Xw_Ahu_Ycl6_716b4ff46c.jpeg?token=fNTbEfuXXkwjgrIX5JqBH5yE4U9XALnZ_ZnwIa_CnnU&amp;expires=1777449427" alt="MPTW2xeaRS1VIdWyDWvkAUzzMc9LShrXwAhuYcl6.jpeg" srcset="https://cms.aihub-oost.nl/uploads/thumbnail_MPTW_2xea_RS_1_V_Id_Wy_D_Wvk_A_Uzz_Mc9_L_Shr_Xw_Ahu_Ycl6_716b4ff46c.jpeg?token=_X2dcHwW7sYZF3kAhexnXmQvw0FuN54p_a7gilkJcgA&amp;expires=1777449427 245w,https://cms.aihub-oost.nl/uploads/small_MPTW_2xea_RS_1_V_Id_Wy_D_Wvk_A_Uzz_Mc9_L_Shr_Xw_Ahu_Ycl6_716b4ff46c.jpeg?token=7ZrrJtyD_Qsgt3yt-TkwCNnD75gAqerypnqj8IZvdBk&amp;expires=1777449427 500w,https://cms.aihub-oost.nl/uploads/medium_MPTW_2xea_RS_1_V_Id_Wy_D_Wvk_A_Uzz_Mc9_L_Shr_Xw_Ahu_Ycl6_716b4ff46c.jpeg?token=4YtLWC4qulU4dHMMdrgIgeb3OIrcGArf2sKdy2uqUa8&amp;expires=1777449427 750w,https://cms.aihub-oost.nl/uploads/large_MPTW_2xea_RS_1_V_Id_Wy_D_Wvk_A_Uzz_Mc9_L_Shr_Xw_Ahu_Ycl6_716b4ff46c.jpeg?token=dzY19MAcnYDSmnCH_II4Y0KFVXiBYx2xMfa8E_ib6s0&amp;expires=1777449427 1000w," sizes="100vw" width="2048" height="1153"></p><p>But first, what exactly does Duflex Mechatronics, based in Elst, Gelderland, do? The company specializes in automation and is the creator of many innovative solutions for industry. “What we’re good at is applying technologies in different situations to address various challenges. An innovation for one sector often becomes a solution for another. For example, we developed a process for manufacturing contact lenses. One of the steps in that process is degreasing the material. Shortly afterward, we applied that same technique to the production of coffee cups. The manufacturer of those cups asked us if we had a solution for peeling paint. Now they degrease the cups using our technique and then apply the paint. The paint actually stays on now.”</p><h2>Helping the Google robot</h2><p>“About five years ago, more and more customers were asking us ‘where they could buy AI,’ without knowing what they were talking about,” Dumoré recalls. “My answer was always: ‘You can’t buy it directly, but you can develop it.’ After all, we humans have to feed every AI system ourselves. Google has been working on this for years, and everyone who ever uses Google helps the company with it.” Dumoré explains how it works: “Everyone is familiar with Google’s ‘robot check.’ When you fill out a form online, you’re usually asked if you’re a robot. For example, you’ll see an image of a car on a road. That image is divided into nine squares, and you have to check which of those squares the car appears in. Based on that, Google has built up a massive library for recognizing objects.”</p><h2>Using automation to take work off your hands</h2><p>“That library is partially available to businesses. We’ve built an object recognition system using images from that library,” says Dumoré. “With it, we can detect 180 objects: phones, backpacks, people, and so on. For us, it’s a showcase to demonstrate to our clients what’s possible with AI. But this technology can take work off the hands of staff at, for example, an airport. It can help customs officials identify bottles of liquid larger than a certain number of centimeters.”</p><h2>Optimizing Processes</h2><p>When the coronavirus outbreak began in the Netherlands, Duflex Mechatronics helped manufacturers such as Auping and VDL build face mask machines. The face masks were desperately needed and, as a result, couldn’t be produced fast enough. “Among other things, we developed an AI solution that allowed manufacturers to see how quickly an operator would respond to a machine malfunction, what actions the operator took, and how long those actions took. This enabled manufacturers to determine whether areas for improvement lay with the machines or with the technicians. And that, in turn, helped to rapidly improve the process, allowing as many face masks as possible to be produced.”</p><h2>Talking Machines</h2><p>This kind of AI application also offers a solution to the labor shortage plaguing the industry, Dumoré explains. “Technology is becoming increasingly complex worldwide, while the number of technical workers is dwindling. And those who are available are becoming more ‘niche’—trained in increasingly specialized fields. A technician who understands the entire machine is hard to find. That’s where AI comes in. Then machines can operate autonomously, and camera systems can identify what can be done better or faster.”<br>Dumoré predicts that the machines of the future will be able to talk. “In all languages,” he says. “This is possible by linking AI technologies together. For example, with ChatGPT. Suppose a machine indicates that it has a hydraulic leak; then, using ChatGPT, the system can tell you what needs to be done to fix it.”</p><h2>Inspiration</h2><p>Dumoré’s greatest source of inspiration is iPhone technology. “The iPhone is far ahead of the industry. I predict that the apps we use on our cell phones day in and day out will raise the bar for the industry. In terms of technical progress, the industry is no longer keeping pace with the consumer world. Take the 20-megapixel camera, for example. This technology is built into an iPhone for 1,000 euros, backed by enormous computing power. In the industry, something comparable still costs at least 15 times as much.”</p><h2><br>Incentives</h2><p>According to the entrepreneur from Gelderland, government incentives are needed to make rapid progress. “There are subsidies for the industry, but they focus on research, materials, and tools, whereas what we really need is personnel. The fact that knowledge is being disseminated through the AI hub is already a good step. That will help the industry move forward. Provided you remain realistic; AI isn’t a magic word—it’s the technical experts who have to devise, program, and feed the systems to make solutions possible. Moreover, if AI becomes a magic bullet, we’ll need to establish strict rules for it; otherwise, technology will run away with us.”</p>

Arno Dumoré, Duflex: "Artificial intelligence is not a magic bullet"

These entrepreneurs have paved the way for you

Meet our frontrunners

wbOgYSbdMK8bSImqrKzXB6ymcejhlc8GgYdTIz2u.jpeg

<p>“Artificial intelligence (AI) is often used as a buzzword in the industry, but I don’t agree with that. You can do amazing things with it, but it’s not a magic bullet,” says Arno Dumoré, owner of Duflex Mechatronics, the company that has been building machines and solutions for various sectors for 25 years. “We’ve been using AI for the past five years. Our goal? To make those applications just as smart as those in an iPhone, if possible. Because, yes, the iPhone is ahead of the industry.”</p>