A conversation between Inspekto CEO Harel Boren and Editor in Chief Brett Brune
Harel, what exactly does Inspekto do? And how did your experience as an Israeli tech community veteran lead you to it?
Our relationships as a veterans of the Israeli tech community with many industrial and investment players in Europe, led my cofounder, Yonatan Hyatt, and me to get acquainted with the huge problem of the lack of autonomous machine vision. This in turn, allowed us to review the needs of quality assurance managers, COOs and the many people involved in myriad manufacturing areas. Inspekto addresses the necessary evil that they regard today as machine vision.
What are the shortcomings of traditional industrial machine vision for the range applications it serves?
Any quality assurance manager on any manufacturing line, in any industry, in any geography, in any tier, and anywhere in the world, has quality assurance challenges on his manufacturing lines, as well as many other challenges best addressed by machine vision as a core technology. The way to treat them today is calling an integrator. And the integrator will choose from thousands of available cameras, thousands of available lenses and lighting apparatus and filters and software packages, etc., and create a solution that will be eventually installed on the manufacturing line in the factory. This is a very tedious process. It costs a lot of money. It requires a lot of attention from plant personnel, and requires a lot of professional services. And at the end of the day, what you get is a solution that is hard-engineered to one particular location on that line. This is a main issue.
You mentioned geography. What is your install base, over which areas of the world?
We first focused mainly on central Europe. For example, Baden-Württemberg, the richest federal state in Germany, includes 52% of the world’s automotive industry, and key to many other industries. That was very simple for us, having an office in Heilbronn and taking a short flight from Tel Aviv. However, those firms all have plants around the world: Mahle, a very early user of ours, has 170-something plants, 42 in Europe. We are now operating in 40% of its plants, and this has created a huge wave with regard to demand from other geographic locations, including China, India, the U.S., Mexico and Portugal. This is true of other users moving to our autonomous machine vision systems, too: firms like Bosch, Federal Mogul, Siemens, Daimler, BSH and Flex.
How does autonomous machine vision as a whole new category overcome the problems you outlined at the beginning?
With autonomous machine vision, a quality assurance manager need not call an integrator. A quality assurance manager, or the internal integrator, can take the autonomous machine vision system out of the box, install it on their line without any integrator and without any knowledge of artificial intelligence or machine vision. He can do this using only a mouse and set it up to work, including everything from box to ops within 45 minutes. This is going on now in brands under which 2,862 plants exist today.
How long did it take previously?
We like to say that we’re 1,000 times faster to install at one-tenth the price. Well, that’s not accurate because 1,000 times is 1,000 hours, which is six weeks. And rarely would you find an integrator-based solution that can be installed in six weeks; it’s usually much longer. In fact, we’re installable, for any range of difficulty, within 30-45 minutes—by the quality assurance manager on his own.
What companies are your chief competitors?
The competition is between us, as the first mover in autonomous machine vision, and machine vision. We’re not going to be alone for long: We believe that in two to three years, some players will catch up and be able to provide an offering in that is well regarded. The key players in machine vision are well known: Keyence, Cognex, Teledyne Dalsa and Omron. But they’re all incarcerated to the ecosystem in which the integrator is king. So you have manufacturers of cameras and manufacturers of lenses and lighting apparatus and filters and software, etc, and the king of the hill is the integrator. The integrator has to put it all together. With affordable and immediate autonomous machine vision systems, we expect the market to expand tremendously. It will attract sizable players whose business models couldn’t be built around professional services. These players will be finally able to step into the machine vision space, with autonomous systems that can be run by the customer.
So this is the main way your product differs from the rest of the market?
Yes, you do it yourself. Unbox it, use it. Just like an Apple device. And if I take the Apple example a step further, when you buy an Apple phone, it’s bare: It only has a few basic icons on it. But if you have an Apple phone and I have an Apple phone and they’re both on the same table, your phone is likely very different from mine because I have a set of applications that suffices and carries my needs and you have a completely different set of applications that carry your needs.
Inspekto S70, our autonomous machine vision system, is very much along these lines. Once you’ve bought it and you install it on your line, within 45 minutes you can choose what need you want to satisfy—quality ensure, archive and trace, sort gauge. It’s up you to choose what set of applications carry the value you want to retrieve.
Harel, what’s a good example of your product eliminating integration and customized developments that characterize machine vision tech today?
Mahle is a good example. Mahle is the number 15 automotive tier-one player in the world today. And we have quite a few installations in Mahle. We also did many beta tests there during the years prior to launch. One specific location I love shows the quickness and effectiveness of the installation. Mahle manufactures a plastic-injected cover for an engine there. And we were able to, within 38 minutes, identify a defect on the line, which convinced the customer then and there how useful this system is. The ROI for that location is achieved in two weeks.
The whole installation took less than 45 minutes. I have had a long array of experiences in the Israeli tech industry, and never before I was I able to take a product and give millions of Euro or millions of dollars in value in 38 minutes.