Automating Wheel Inspection with AI
MVPro Media – The Vision Podcast #29
Guest – Philipp Franz, Key Account Manager| Eines Vision Systems
Philipp Franz, Key Account Manager at EINES, joins Josh Eastburn to discuss how AI-powered machine vision is transforming quality inspection in automotive manufacturing, replacing manual visual checks with automated, data-driven inspection systems.
In this episode, Philippe explains how the EINES Rim Surface Inspector (ERSI) automates wheel inspection using machine vision, lighting, and AI to improve defect detection, consistency, and production efficiency. The conversation explores the challenges of inspecting different wheel designs and finishes, integrating inspection data into Industry 4.0 manufacturing environments, training AI for new product models, and how automated quality inspection is helping automotive manufacturers reduce defects while gaining deeper insight into their production processes.
On this page:
- Podcast player
- Sponsor information
- Guest information
- Useful links
- Episode chapters
- Episode transcript
Listen to the Episode
Episode Sponsor
About our Guest
Philipp Franz is Key Account Manager at EINES Vision Systems, where he works with automotive manufacturers and Tier 1 suppliers to implement AI-powered machine vision systems for quality inspection and production automation. With a background in mechanical engineering, Philipp specialises in helping customers improve manufacturing quality through automated inspection, defect detection, and Industry 4.0 integration. At EINES, he supports the deployment of advanced vision solutions across the automotive production process, from wheel and tyre inspection to final vehicle quality control.
Useful Links:
Philipp Franz: franz.philipp@eines.com | linkedin.com/in/philipp-franz-automotive-quality-control/
Eines Vision Systems: eines.com | comunicacion@eines.com | linkedin.com/company/eines-quality-engineering-vision-systems | youtube.com/@einesvisionsystems4159
Webinar – The (R)Evolution of Wheel, Rim & Tire Quality Control: eines.com/event/the-revolution-of-wheel-rim-tire-quality-control
Episode Chapters
Click onto the chapters to access the relevant sections of the transcript below.
1. From Manual Inspection to AI-Powered Quality Control
Philipp explains the limitations of manual rim inspection and how automated machine vision systems improve consistency, accuracy, and defect detection on automotive production lines.
2. Designing a Rim Inspection System
The discussion explores how EINES configures inspection systems for different wheel models, covering camera placement, lighting, installation, AI training, and the factors that influence system design.
3. Using Inspection Data to Improve Manufacturing
Philipp explains how inspection data integrates with existing manufacturing systems, providing traceability, production analytics, and feedback that helps manufacturers identify recurring quality issues and optimise production.
4. Expanding Automated Inspection Across Automotive Production
The conversation concludes with a look at EINES’ wider portfolio of machine vision systems, including solutions for tyre inspection, body shops, paint inspection, gap and flush measurement, and final vehicle quality control.
“Humans generally are not the best at repetitive tasks… a machine doesn’t care about this. It’s always giving the same effort.”
Episode Transcript
Philipp Franz – Eines Vision Systems
We’re able to tell basically in which kind of models or which color, in which area of the rim have more defects, have less defects. All the data that is collected automatically and we have a feedback loop with a cycle time of like 6 seconds.
Josh Eastburn – Host
Welcome to the MVPRO Podcast. Way back in the early days in wee little episode number 7, we talked about some of the challenges of automotive inspection systems with EINES VISION Systems, a KONICA MINOLTA company. Today, some 20 episodes later, we are talking automotive with EINES again, although you will see we have a discrepancy with the pronunciation this time because my guest is located in the Spanish headquarters of the company where the name is pronounced EINIS. Don’t get confused. For more than 30 years, EINES has helped manufacturers improve quality control through automated inspection systems, deep learning technologies, metrology solutions, and defect detection systems. Its solutions are used by automotive OEMs and Tier 1 suppliers worldwide to improve efficiency, reduce defects, and support smarter manufacturing processes. My guest today is Philipp Franz, Key Account Manager at EINES, working with automotive manufacturers and suppliers on machine vision and AI-based quality inspection solutions. With a technical background in mechanical engineering, his focus is on supporting customers in improving production quality, efficiency, and automation through advanced inspection technologies. Today we talk about EINES’s recently launched system for automated rim inspection called ERSI, ERSI, or ERSI, as you’ll see.
Josh Eastburn – Host
But first, for MVPro Media subscribers, look out for VDC Strategies’ Machine Vision Investment Survey in this week’s newsletter and share your views on the priorities shaping the future of the industry. VDC has tracked machine vision markets for over a decade, and your input will help shape this year’s benchmark study. Participants will receive a complimentary summary of key findings and be entered into a drawing to win 1 of 10 Amazon gift cards. If you aren’t signed up for the MVPRO Weekly newsletter, head over to mvpromedia.com and click subscribe. Okay, let’s get into the interview.
Josh Eastburn – Host
How did you end up living in Spain? What did you come there for, EINES?
Philipp Franz – Eines Vision Systems
No, it was in 2024. I was still doing my mechanical engineering studies. And the thing is, in my bachelor degree, you could do like one full year abroad. So yeah, I applied for it, then I got here, and then I liked it so much, and I finished my studies here, started working, and then I found a job here at EINES. It is called EIN-ES, by the way, not EINS or EINIS. Yeah.
Josh Eastburn – Host
You said you were studying mechanical engineering and you had done some design work. —before coming to work for EINES. Exactly, yes. Had you had any exposure to machine vision before starting in this position?
Philipp Franz – Eines Vision Systems
Not really a lot. I was working for companies that produce and develop lights, but I was there more in like a mechanical department, then also for company Stadler, alsovery well known, Stadler Rail, more mechanical engineering department and not really a lot of machine vision.
Josh Eastburn – Host
Okay, so what were some of your first impressions of the technology when you started at at EINES.
Philipp Franz – Eines Vision Systems
My question was always like, how does it actually work? Like, it was, it’s very, very interesting how some things can work with the machine vision, with light, with cameras and so on. And then you just have AI working for you and all the possibilities that you have there.It’s, it’s huge. So it was for me like very interesting. Also for me, it was a great opportunity. And of course, because it is for automotive as well, it is for me also personally very interesting because I’m also a fan of cars.
Josh Eastburn – Host
So let’s talk about ERSI specifically. The full name of the product, correct me if I’m wrong, is the EINES Rim Surface Inspector.
Philipp Franz – Eines Vision Systems
Is that correct? EINES Rim Surface Inspector. Exactly. We also call it ERSI, the ERSI.
Josh Eastburn – Host
So for someone who hasn’t seen that on a production line, can you describe what the system is actually doing step by step when a wheel comes through and how does that compare to how rim inspection is handled now?
Philipp Franz – Eines Vision Systems
Let’s start first of all with the manual inspection, how it is usually done nowadays. So the manual inspection, of course, let’s imagine you have a, you’re a car manufacturer and you have your supplier for the rims and these rims, they get produced, they get forged or whatsoever. And you need to have a manual inspection in the line, or let’s say a general inspection to ensure the quality. And this manual inspection is by having some people, some workers on the line and only using their own eyes to see, okay, here we have defect here on the rim, is a bigger one, a smaller one, and so on. And if this is done by humans, there’s always a, like, a little bit of a problem in there because humans generally are not the best, I would say, at repetitive tasks. Because it’s more or less always the same. You look always for the same. And by doing this for, let’s imagine, for a whole day, for 8 hours, you do this, sooner or later you will get tired. Of course, we have our own weaknesses, I would call it, like normal things as a human is, and you probably miss a defect or you don’t.
Philipp Franz – Eines Vision Systems
So this is exactly the point where then the automated inspection, that’s now the ERSI for rims, comes in place where we replace this manual inspection with an automated inspection. So how does it work? We have two different options for the ERSI, inline or offline. Let’s go with the inline, for example, where we have our— it’s like a little bit of a box that you place on top of the conveyor where the rims pass through the rolls on the conveyor or on the belt, and it gets placed inside of this box. Okay. And with a cycle time of 6 seconds, something like that, we, um, fix it on a specific point and make photos of it together with lights. And then together with our software that is also from us, we detect all the defects. And by doing this with AI together, you have an automatic inspection. So you get rid of all the manual inspection of the eyes of the humans. And replace them with cameras and lights. And that’s how you have a much higher detection rate. The defects that you can see or can detect are much smaller. You are more reliable. And again, regarding repetitive tasks, a machine doesn’t care about this.
Philipp Franz – Eines Vision Systems
You know, it’s always, it’s giving the same effort. You know, a machine is not having this kind of problem. So that’s how the machine works. And yeah, so to explain the machine maybe a little bit more. It gets placed inside the box, gets rotated, and photos are made. And these photos get captured, get saved locally. And then with our software, we’re able to say where these defects are and what kind of defects they are. But to add this as well, it can also be offline where a worker takes this rim and places it inside the box, presses a button, and then it gets inspected. But this is then not a fully automatics, process. This is more like a manual, but still with the inspections automatic.
Josh Eastburn – Host
Gotcha. Okay.And a cycle time of about 6 seconds, that’s impressive to capture all of that information.
Philipp Franz – Eines Vision Systems
Exactly. Of course, it’s very important because time is always money.
Josh Eastburn – Host
So on the website for the product page for ERSI, it advertises an installation time of just a few days, right? Does that installation time, in your experience, include the time to train the system as well, or is there a real onboarding time in addition to installation? That it takes in order for the system to understand how to recognize all these defects?
Philipp Franz – Eines Vision Systems
We develop it and we do it here in Valencia, in Spain, at the offices here and we then give it to the customer, we ship it to the customer, and the, the installation then there will take a few days, maybe a weekend. Okay. So, and then afterwards the production can still be ongoing, but we then need for, we need a specific time, of course, to train a specific model and could be of, let’s say, for one model of around one week, something like this, until it’s like fully developed and can work for it, you know? So this is something we can do as well, like as the company, but it’s also possible that we offer trainings, for example. So they’re always able to do it for themselves in the future when a new model is implemented. The general machine is developed, is fully developed. And, but of course you always need to have some adaptations depending on the amount of models that you have. If you have more models, you would need more cameras or more light, less light. Depends of course on the specification, the needs of the client.
Philipp Franz – Eines Vision Systems
So every machine is a little bit tailor-made.When you have different size, the cameras, let’s say for example, you have one camera which is for one model, which has specific size. Then you have a different model, which is bigger. You need You probably cannot see everything with the camera because of the different size. So you maybe need to add another camera to see the whole surface. And that, of course, the more models you have, or the more you want to see, when you want to see something from the bottom as well, from the side, the number of cameras changes every single time. Okay. But we also do like simulations. We also do these kind of simulations beforehand to tell the customer then in the end, you what we can see, what we cannot, and before even installing the system or fully developing and manufacturing the system, how many cameras we would need by doing a simulation.
Josh Eastburn – Host
So part of the upfront design process is they’re talking with you about how many different types of models, what sizes of rims that they need this system to, to inspect.
Philipp Franz – Eines Vision Systems
Exactly. Very, very crucial information. Yeah.
Josh Eastburn – Host
Other details like that, that are, that are kind of crucial to that design process?
Philipp Franz – Eines Vision Systems
Yeah, of course. The size and so on, how the, how the facility looks like, what kind of conveyor, the speed and so on. And of course the minimum size defect, for example, that you have, that you want to detect, what kind of defects, what kind of colors, materials, and so on. How the facility looks like, how much room do you have available? These just to name a few.
Josh Eastburn – Host
Okay. So all of that is taken into consideration in delivering the, the final design.
Philipp Franz – Eines Vision Systems
Yeah, of course.
Josh Eastburn – Host
Okay. And then let’s say, let’s say a, A system is now an, an ERSI system is installed and in place and operating, and the end user wants to introduce a new model to that automated inspection line. Do you go through that whole process again or, or does that require bringing in a new ERSI system designed to work with that model?
Philipp Franz – Eines Vision Systems
No, you optimize the one you have already because the software just maybe needs an update. You maybe need another camera. And you need to learn, of course, the new model with AI. That’s it.
Josh Eastburn – Host
Okay. So the, oh, and then they, if it involved installing new lighting, new cameras, the Ines team would come out and make those adjustments.
Philipp Franz – Eines Vision Systems
This is one option. Yes. Or we have that option, as I mentioned already, we offer trainings as well, which we offer to the customer. They’re able to do it themselves. We also offer like customer success packages, like some maintenance packages. We also like, for example, like 24/7 availability to offer customer support, also remote support as well. So we are also always available to help this. We also travel a lot, like we as a technical department or me myself as well. And therefore we have many different kinds of options.
Josh Eastburn – Host
EINES on the product page for ERSI describes it as being able to integrate with Industry 4.0 environments. So new smart manufacturing types of systems, and specifically for things like real-time traceability, defect reporting, production analytics. In practice, how does ERSI integrate with an MES, SCADA, quality control systems that might be in place in a facility already?
Philipp Franz – Eines Vision Systems
Well, first of all, like to connect with the programs that are already used from the customer, we are able to connect with everything. There’s a big flexibility in that. And then how we connect it, of course, the thing is, all the data that is collected, all the pictures, the data of the defects that are captured, we store them in the client’s cloud, and the client can do whatever he wants with it, you know, so we offer a big feedback, all the data that is captured, and in the end, can be used from the customer. And we are able to tell basically in which kind of models or in which color, in which area of the rim or whatsoever have more defects, have less defects. And we provide this information. Yeah. And this information can then be used to make the manufacturing process of the rims smarter by telling, okay, in this area we need additional investigations because we can see, for example, okay, in this area we have like always a deep defect, always this kind of defect, but why? So this is all the data that is collected automatically, and that’s how we have, like, let’s imagine, like, a feedback loop.
Philipp Franz – Eines Vision Systems
So let’s just make in the manufacturing process with an arrow, we go back to the start, and while going back, we have this information where we need to have an additional look at, and that’s how we improve the system.
Josh Eastburn – Host
Are there any, like, let’s say, what would be the, maybe the most challenging aspects of room inspection that ERSI is designed to deal with? Is that material? Is it dark materials, reflective materials? Obviously we can have different types of alloys and things that we’re looking for, material defects, right? Porosity or something like that. Yeah. What’s the worst case scenario that you might have to, that, that is acceptable for the ERSI system?
Philipp Franz – Eines Vision Systems
What I wanna say is to go to very small defects, of course, the smaller they are, the more difficult it gets. But in the end, it always depends what the customer has or what he wants. Some customers have more strict requirements, others have less strict requirements, like materials or colors or whatsoever. This is not really a problem for us. We can detect more or less like everything. Yeah, I would definitely say like one of the biggest is always like with the defect size itself and how the rims are shaped to reach, let’s say for example, 99% detection rate is always like difficult. We always try our best for each one. Um, but yeah, it always depends what kind of models do we have, how many different. Let’s say, for example, we have hundreds of different models. It’s always also difficult to capture all of these in one machine. So to name a few things.
Josh Eastburn – Host
Is there ever a situation where you need to do, let’s say, multiple passes, or, or maybe does it need to spend a longer time in, in in cycle, in the inspection cycle?
Philipp Franz – Eines Vision Systems
Cycle time isn’t really changing in the end. You just put more cameras in or more light, but this can then be, of course, challenging to detect every little defect and every different kinds of defects. That is always then challenging with a lot of different models then, yeah.
Josh Eastburn – Host
Mm-hmm. Are there any cases where you wouldn’t recommend that a customer use the ERSI system instead of a manual inspection?
Philipp Franz – Eines Vision Systems
I would always recommend it because we could always make like a simplified version of it. If I was just thinking of, let’s say for example, you had, you don’t have a big volume of it and you just want an easy inspection, we could also adapt our machine. We could also simply adapt our machine to that, could do less, make a simpler version of it to still get some kind of defect detection. So I would always recommend it. It is always a very great opportunity to improve the process.
Josh Eastburn – Host
If I’m a, you know, manufacturer, quality engineer who wants to understand where ERSI might fit into my operation, what’s the first step I should take for getting in touch with your team?
Philipp Franz – Eines Vision Systems
I would say directly with me. Somebody can like email me directly and what I would like to do, of course, is first of all to have a little bit of call of a Teams meeting, of online Teams meeting to first discuss a little bit of the use case, some general aspects, and then furthermore, we go deeper into the details. Otherwise, when, as I said, when we do it with WhatsApp, for example, or with EINES mail. In the end, it will land at my desk and we go from there.
Josh Eastburn – Host
Mm-hmm. Is it possible to see a demo of the ERSI system in operation?
Philipp Franz – Eines Vision Systems
A demo itself? What, what we offer as well is that companies, clients visit us here in Valencia in our offices. If we have an interested client who has a project going on, of course. We offer that they come and visit us as well, and we would like to show the machine here in our office, in our showroom, in our big showroom here. There are also some videos that we have uploaded on LinkedIn. So LinkedIn is also a platform that we use a lot where the marketing department is doing their work. Definitely, I want to say, as well as on YouTube. So we also have our own YouTube channel where we upload some videos. There are also videos available for the ERSI system.
Josh Eastburn – Host
Perfect. I saw some of that and it’s pretty impressive. Outside of ERSI, is there anything coming up at EINES product releases, events that people might want to pay attention to?
Philipp Franz – Eines Vision Systems
The ERSI itself is one of, is our newest solution in that case. So we’re definitely focusing on that to improve that system. Also for the tire inspection and then also we have further systems regarding wheel and tire inspection. But generally all of our products, as you might know as well, we have like products for every point in the manufacturing process, in the automotive manufacturing process. We also offer sometimes some webinars, for example, we just recently, I think 2 weeks ago, we had a webinar where we had a, I think it was a full hour presentation of the ERSI system as well.
Philipp Franz – Eines Vision Systems
Okay, great. I’ll follow up and ask you for links to all of that. We can share the link to the webinar, to your YouTube channels and whatnot, and then, yeah, people can follow up that way too. Fantastic. Are you looking for, I know summer is hot in Spain. I know it gets very hot. You’re next to the beach, but I know it’s hot. How’s it feeling right now?
Philipp Franz – Eines Vision Systems
Well, the weather, of course, Spain, Spanish weather, it is always like 30, 35 degrees right now at the moment. So it can get very hot. But of course, generally the beach helps.
Josh Eastburn – Host
Our listeners are going to be jealous.
Philipp Franz – Eines Vision Systems
But of course it is.
Josh Eastburn – Host
Yeah
Philipp Franz – Eines Vision Systems
I can recommend 100% living in Spain.
Josh Eastburn – Host
Awesome. Was there anything else that you wanted to talk about that we didn’t discuss already?
Philipp Franz – Eines Vision Systems
So if anybody’s like interested in connecting with me, the quickest way is to connect with me via mail. So we have franz.philipp@eines.com. Feel free to contact me and we can go from there. Otherwise, I just want to put in the information as well that we have solutions and products for the whole manufacturing process. So not only for the wheel inspection, but also, for example, for the paint shop, also for the body shop, where we also inspect stamped parts as well, profiles. Then to go further in the final assembly as well, where we have inspection tunnels as well for the full inspection for gap and flush, for example. Then we also have for the final assembly on, on surface damage. Inspector for scratches and dents in the final assembly. Also like spec check tunnels where we have a look at, let’s say for example, the correct logo is placed, the correct wheels, the correct light and so on. So we have a very broad solution portfolio and therefore I’m sure we can have the correct solution for any kinds of real, the claims in the end.
Josh Eastburn – Host
Wow. Yeah, that’s very complete. Well, thank you for your time today, Philipp.
Josh Eastburn – Host
While we’re on our way out, let me remind you to open up the MV Pro newsletter this week and look out for VDC Strategies’ Machine Vision Investment Survey. They wanna know what you are prioritizing and they are staking some sweet, sweet gift cards on that opinion. So go respond to the survey and get one for yourself. Either way, all participants will receive a complimentary summary of key findings once the report is published.
Josh Eastburn – Host
That was Philipp Franz of Eines Vision Systems. To see more quality inspection solutions for the automotive industry, visit eines.com. And if you’re working in automotive vision inspection or any of the many, many other fields applying machine vision every day, and you wanna share your story, your expertise, or your insight, please find me on LinkedIn or at josh.eastburn@mvpromedia.com. This episode was produced by Flanur Creative Studio and Big Robo. For MVProMedia, I’m Josh Eastburn. Be well.
