The Manufacturing Executive
The Manufacturing Executive

Episode · 5 months ago

Custom Machine Vision & Robotics Solutions w/ Jonathan Berte


Mass production alone is not enough. We all want customized products.

Twenty years ago, it was easy. Educate your workers about your single unique product. In a world driven to customization, however, no worker can learn 400 different products. No *human* worker, that is.

But a digital worker might. Could robotic camera vision streams be the solution?

In today's episode, I talk about building custom machine vision and robotics solutions with Jonathan Berte, Founder at

Here's what Jonathan and I discussed:

  1. His personal journey into the vision component of robotics
  2. How AI's vision is advancing right now
  3. The most important applications of AI in the manufacturing process

To ensure that you never miss an episode of The Manufacturing Executive, subscribe on Apple Podcasts, or Spotify, or here.

Listening on a desktop & can’t see the links? Just search for The Manufacturing Executive in your favorite podcast player.

That this moment, mass production aloneis not enough. We all want customized products, and this customization isjust hiring the manufacturing complexity and you can just edicate TAworker. I would have variety of four hundred different types of products atwere just one unique type twenty years ago. So it's just much more easy to putthese flows into robotic camera vision, streams, welcome to the manufacturing executivepodcast, where we explore the strategies and experiences that aredriving midsize manufacturers forward here. You'll discover new insights frompassionate manufacturing leaders who have compelling stories to share abouttheir successes and struggles and youill learn from BTO B sales andmarketing experts about how to apply actionable business developmentstrategies inside your business. Let's get into the show, welcome to another episode of theManufacturing Executive Podcast, I'm Joe Sullivan your host and a Co founderof the Industrial Marketing Agency Gerrilla. Seventy six. If you ask myfour year old son Jack about robots, he'll go running for his optimist,prime figure or one of the other transformers that are scattered acrossmy house. By the end of every day. When I hear the word robot, I buy default.Think of big, yellow, robotic arms on an assembly line moving a hundred milesan hour. I guess today, however, is an expert in a different part of theindustrial robot and that's the eyes. So let me take a moment to introducehim. Jonathan Bertay is the mastermine and driving force behind Robo Vision,the collaborative platform for those looking to lead the artificialintelligence or AI revolution, with a background, as an applied physics,engineer specialized in image, processing, Jonathan built up a strongreputation, building custom machine vision and robotic solutions in thefirst years of Robo Vision's existence, the robot, Robo Vision, Ai software, iswhat it is today and large part due to Jonathan Building Machine Vision,applications the hard way Jonathan now passionately leads Robo Vision to newheights every day. Jonathan Welcome to the show. Thank you joe well go glad tobe here, yeah, it's great to have you. This is such an interesting topic, andyou know I talk to a lot of people who touch robotics and automation from avariety of perspectives, but you're kind of in this. This corner a relateddovision and you got some deep expertise there and I'm I'm verycurious to learn from you and sort of expand. My own knowledge, as as youtalk about you, know your expertise. Okay, so you I guess like, like, I said,Jonathan to kick it off here. I know you're in this world of vision, vision,component of Ai and I'm just curious if you could start by telling our audiencea little bit about your personal journey that led you into that nicheand then inspired you to found Robo...

Vision. Well, it's a good question showit all started about twenty five years ago, when I started experimenting withGamerous and when I realized that is, migration from analyc cameras todigital cameras had also some a great potential, and I started basicallydreaming about the potential when I was at University of Syric Switzerland yearprogramming abot. That could look behind the back of a Croupier in acasino so just to follow the game, and that's where I realized that not onlythese gimiky applications wourds be cool, but also image, processing andeverything related to getting value from image. streems cameras also inproduction conditions witd at some point in time, be very desruptive and Ibelieve that early on I started slowly as a consultant and the outomotivemarket. I was really the the careful first steps. We all know that theindustrial space is very conservative, very robustes, driven typical componentlifetimes of five to seven years, so it was not an easy GAM in the beginning,especially because of plcs Ye had e standard as seven semens Allen, Bratleythe usual suspects Omron being there, and this camera ecosystem was kind of anew kid on the block, but also a difficult new ket on the block, becauseit was always the last component in an automatian project and no stricky one,because you didn't have the production data. Yet so that's where I felt thehard way. WITD industrial conditions are backe. Then I really had a toughtime, those first years, two thousand and four two thousand and five andconvincing people that there could be potentiall in this camera application,but that there was simply not so much support. There was no were no large codbases out there th the was the term ai was, was really not used at a time andit was seen as an extra component to a peal c. So every budget was like toomuch and can to you reduce the budget, somehow with less features, but in theend it was a very challenging environment that until one points whenthings started to change when digital cameras, game more and more comoditizedand and this whole a I revolution- basically starting oven at Standfort,California, with Fai Ale, imagenet started to take off, and people startedtalking about deep learning and that's when I deeply understood that thebiggest dinstruption of always not the camera itself, looking at productionconditions and in quality control production line. Now it was thealgorithm creation. It was around two thousand and ten to thouand eleven.When I basically started all over again and said, like my business models, notscalable, I need to do something about...

...the algorithm creation itself andthat's when I went back to university personal friend of mine, that wasprofessor at a time, a really cool machine learning lap that he had it andat a time some cool new guys were in this lap, Sener dealeman and AranVander port that are now like high profile De Mint Engineers, and theyexplained me- that's writing algarithmicsofter, who was coming to anend with the advent of deep learning and that's why around two thousand andthirteen I really started off with Ro Vision, two point: Zero. We had a, Imean designing a platform that made algorithm creation, intevisual domainautomatic and that's Yeh, that's the basis of f the company. Basically likeyou can create algorithms, whether it's quality control and the industry orGovi detection in healthcare. You can automate this process, which is thebiggest disruption of Oljo. That's a really great background. T S. it's suchan interesting technological advancement. I'm just curious! If youcould talk a little bit about you know some of the specifics of how the visioncomponent of Ai is advancing right now and like what's possible now that maynot have been possible even a few years ago. Well, basically, if you look atthe vision problem, you typically have this aspect of an object or somethingthat you need to detect or some some aspects or artifacts that you need todetect from the object, and you have background and in the past, likefifteen years to ten to fifteen years ago, you had to engineer all of thisbackground. substraction for ground extraction, feature, detection andrebustess. You had to do that all by yourself, meaning that it was reallyyou you needed to hand pick the features you needed to know all aboutthe possibilities. So typically a profele that had to write such anAugorithm was already like PhD degree. I mean you needed like a pgd to besuccessful and then to create a robust image processing augarden fifteen yearsago now this changed when researchers started to reverse engineer the visualPatway in the brain, and they realized that they started to realize some ofthe basic aspects of how a biological brain a human brain is processing,visual information and basically that was the advent of deep learning whenactually the brain people started to reverseengineering, the brain and- and this feature, detection that was like amanual dot fiften years ago, became autimated, and it's, of course, a biggame changer, because those are the most expensive profiles in the industryis the algorithm creators. So if you can just autimate the smartest people,all of a sudden, you can just do many...

...more use cases and make many morealgorithms in that's the era we live in, it's the era of the commode desation ofdifficult algorithms, especially in the visual demain. What are some of themost important applications of AI in the manufacturing process? From yourperspective, Jonathan and also I was curious if you could maybe share a reallife example, whether it's from Robo Vision or, if you you know from from acustomer or just something tangible, to put some of this in context forlisteners, we all wear clothes like catdrousers yeah anything. So these are processes that are goingextremely fast, but if something goes wrong with the needles- and you havethis textile being woven and very high speeds, if, if one needle breaks orsomething- and if it's not like noticed early enough, you have like literallyhundreds and hundreds of meters who wasted tex stile really needs to go. Imean down the drain, and this is where a I can be of Atermostixcan can reallyfunction, wellit, it's like to detect hie anomalies in involving dextyle andas such, stap the production process, if a needle breaks so not to waste anymore material. That's a great example, so quality control quality assurancewould be sort of one place and anywhere else that you'd you'd. You touch onobject segmentation, like everything related to detecting an object, thepose of an object. For instance, a glue process which we did for theaudefactory in Brussels, like many parts of a car being glued withhigh tech, glues that then, of course, you need to check if, if the glue hasbeen put well on material and this kind of of processes can be ultimated to itsAI aneffisioal way, but also picking like object. Picking, if you have likea convayer belt, whel of product needs to be packaged or need to be qualitycontrol before packaging. Then you, you typically need to know the position andthe exact post, or you can pick it up with a robot and put it in a box, andthat's great, you know one common theme that seems to emerge from week to week.In my conversations with manufacturing leaders who are coming from alldifferent kinds of backgrounds in the industrial sector is this concept ofthe manufacturing labor shortage or this problem I should say, and when youand I talked a few weeks back an preparation for this discussion, yousaid something along the following lines: I'm probably going Ta Butcheryour quote, but is how I wrote it down. You said something like if you want todouble revenue, finding the Laboris, the biggest barrier and I'm just kindof curious to hear you talk a little bit about what you mean by that and howyou see the role of Ai, alleviating that pain. Well, many of the, forinstance. Quality control issues are...

...very tedious tasks and they need to bedone literally twenty four hours a day. So at four in the night you are aworker and you need to look at textile for hours in a row with big lampsshining in your face. It's just simply not going to work. So many of thesepeople get tired. They get it's difficult to keep focused. So at somepoint you will just waste material because it's not what humans aresupposed to do like looking at textile hours in a row. So if you don't havethese profiles or if you cannot automate this process, then you willjust just have more material ways, so you will have revenue yeah. You havemore costs that and also these profiles, especially in very industrializedcountriess, like Belgium and Hollands they're difficult to find. So you havethis typical labor migration aspects: People coming from Eastern Europe inour region- that's yeah, it's just very difficult to keep production runningwith Labor shortes, especially in pandemic times, when you have all ofthese regulations keeping distance. So it's it's a whole challenge to keep thefactory running at the heat of the epademic and are you finding from you ntalking to your own customers and other people in the manufacturing space thatthe problem around Labor is that people just don't want to do those jobsexactly exactly Yeah Wbel? What you have is, I don't know if you havechildren ive tree, that one of the funniest movies they like to watch this.It's just like. I mean it's decades old, even more than hundred years old. Isuppose it's Charlie Japplin and we all nothice this image of Charlie, comingback from work and still having this kind of movement in his arms because hehas been screwing all day, the same screw a d and our children just laughwith it and they say like. Oh that's just so funny. It was this like reallyhappening, or is this just a movie? No, it's really happening, and still thereare a lot of jobs that are just very been pointing one task. I respect alloperators in ourl backgrounds, but there are plenty of other jobs to dudeand then just yes, screwing one screw. That's just not a very humane thing todo forty hours a week, yeah! That's that's a and I love your charleteCharlie Chaplin example. There it's you know because it's it is true, and it'sand everything that I'm hearing is that you know the younger generation cominginto manufacturing that they just don't want to do these jobs. You know the thebaby boomers are exiting the workforce, the ones who who ere comfortable doingthese jobs, maybe more than than the younger generations and so who's goingto do them. And you know the reality is: There's a big opportunity and roboticsto fill jobs that people frankly just don't want to do and as a result, maybefactores are kind of left. You know stumped with what. How do we fill thefill the void so make sense, and that...

...then also the the aspect of more andmore customization like industrial revolution, has of course come up withmass production. But at this moment mass production alone is not enough. Weall want customized products, and this customization is just hiring themanufacturing complexity and you can just edicate TA worker. I would havevariety of four hundred different types of products at were just one uniquetype twenty years ago. So it's just much more easy to put these flows intorobotic camera vision streams. That's great one takeaway! I have hereis I'mgoing to have to see if I can get my four year old and six yearold watchus,some Charlie Chapon, instead of just po patrol, so well work on that Jonathan I've heard you talk a littlebit about cloud versus edge technology as it relates to scaling your qualityoperations, and this is, admittedly, a knowledge gap. For me, I don't know awhole lot about this and differentiating between cloud and edgetechnology, they're kind of buzz words to me, and I'm wondering if you can enlighten meand maybe some of our listeners, who maybe feeling the same thing about sortof where this fits into the discussion that we're having here today was a verygood question, and some years ago cloud was seeme contradictory to productionmanagement systems and quality control is system because everybody wasresponding as along the same lines like what, if you're into Internetconnection, just false breaks down, and, of course, that is, that can still be apossibility. But with five ce with his low latency networks, high band withwith not only a fiber connected to your company, but also yeah other kind ofsecond choice or packup systems. Cloud is really becoming of primorialimportance in quality control. Because of the fact that you can just easilyscale up. You can just have yea the scaling operations run by your cloudprovider in contradiction. If you have like an EG application, if you want to have have double production or a createa new production line, you literally need to go and buy components. You needto have this. This edge devices installs and calibrated, so thiscalability is less with edge, but in some uwsecases and we have industrialcustomers that work at Ar dapt environment, also because of production,secrecy reasons and so on. Then you of course need to resort to de compute andDGE. Compute is basically a a I hardwore defice near to the productionas side. That is just having a very small loop image, acquisition andfearings of the AMO and obrational results that will lead to a PLC likepushing out a product if it's below quality so t. That is basically the thereasoning between cloud versus Gh. So cloud is your computes localized at aclod provider, asyour Microsoft, to a...

Google cloud or Amazon web services,whereas your local edge is device such as an Nvideo jetson like a small GPU,which is very near to the production, where it happens, some cases, evenabove the Prov that convayer belt. Looking at the product and and giving adirect outpit to the PLC to steer some operation so both have you used cases.Clad is more scalable, it's also easier to update, because, basically it's justyour cibe provider updating to nugpws or something so the game is verydepending on the usecase Joe. That's great. I appreciate you kind of givenme some more insight into that topic. Well, Jonathan! You, you and I sort ofmet via the industry. Four Point: Oh Group, that is sort of taking root onclubhouse right now I keep talking about clubhouse. I feel like in everyevery episode of this podcast, because it's just becoming now a reallyinteresting place and I think the some pretty cool things happening inthe manufacturing sector. There I was, I thought I just kind tof, ask you whatwhat do you see in on that platform in terms of conversations that arehappening around ai and robotics and just kind of curious? How that's sortof influencing you and the potential you see there? I just love it becauseof the direct correct. You can really make a connection with yeah, some other,a IENPRENEUR or somebody responsible for production and go into thesconversations. It's a really good tool to to learn about the world. How otherpeople and other teams are solving problems in other continents to sinkand give each other advice. I think it's the next big thing in social media,but I'm biased, because I'm kind of a bit addicted to clovehouse right now. Ican understand how it would happend it's that I've participated in as an observer, largely in thatparticular group industry for point O, mostly just to sort of see what peopleare talking about and learn and enhance my understanding of some of thesetopics. And then I've been you know, sort of running some of my ownconversations more related to marketing and sales and in the manufacturingspace. But it's pretty interesting, there's a lot of smart people. You knowit's. I don't want to say: I'm surprised that the manufacturing sectoris there, but I am a little surprised at how much it's taking root and socurious to see kind of wer where things go on that front. Yeah and also to formnew relationship to talk to potential partners, it's just so much more easythan a cold email, just to have a conversation just yet talking formaland afterwards about. It is very specific brooging, something for suretohe, actually hearing someone's voice, rather than just reading words on thescreen. Right, yes, great! Well, Jonathan! Is there anything that I didnot ask you today that you Yu'd like to touch on while I've got you here. NotReally I mean we didn't really. A lot... changing also in the in NorthAmerica, with Poston dynamics and all of these new robotic frameworks Berkleybased. So I do think that flexible manufacturing is really taking off ifrubolts are much more easy to program and with, if you combine the learningwit advanced robotlics, I see yeah a lot of difficult dusks like welding andlike moulding and CNCA manufacturing being replaced by smart robotics yeah.It's such an interesting space to be watching right now recently on one ofthese episodes ahead, Ryan Lillibridge from mission design and automation onhere, and he was w. He brought up the fact that robots have been around nowfor something like sixty years, which is which kind of I almost had ask me.Did you say sixty or do you mean twenty oar? Thirty, because you know it, youalmost don't realize it, but some of the advancements that are happeningright now are really kind of t, there's just so many cutting edge things. Imean you've talked quite a bit here today about you know where vision fitsinto that, but you know from robots as a service to being able to you know theadvance men, ips and cobots and robots working alongside people in safeenvironment, taare becoming more are safe, whereas they, you know the robots,ere, always traditionally caged, and you know, I think it's just it's reallyinteresting how fast things are changing absolutely, and I yeah I'mreally looking forward to this future, where robots are even becoming moresmarter than they are today very passionate about thechologicaladvancements, great well, Jonathan. This is a really great conversation. Iappreciate you doing this today. You welcome Jo, was really fun to do great,and can you tell our audience about where they can get in touch with youand also where they can learn more about Robo Vision, yeah the best way toget in touch with me is Vio Linton, just johntan Bertie and the youll findmy profile or to follow the Robo Vision, dot a I linkedon page or just a websitethrough provision. Alt Ai see what usecases we have, how easy it is tocreate vew augorithms with our platform and how we can solve your nextchallenge in manufacturing. So itwas great talking on this platform, I'mreally looking forward and bringing to you some lessons learned that we'velearned in manufacturing halls here with a highropotics in Europe and openthe American market for advanced a I technology. Great Will Jonathan onceagain really appreciate you doing this today and as for the rest of you, Ihope to catch you on the next episode of the Manufacturing Executive. You've been listening to themanufacturing executive podcast to ensure that you never missed an episodesubscribe to the show in your favorite podcast player. If you'd like to learnmore about industrial marketing and... strategy, you'll find an everexpanding collection of articles, videos guides and tools specificallyfor be to B manufacturers at grilla. Seventy SIXCOM flashalarn. Thank you somuch for listening until next time.

In-Stream Audio Search


Search across all episodes within this podcast

Episodes (70)