TECH B2B Manufacturing Matters Featuring Vision Markets

Winn Hardin, TECH B2B Marketing:

Welcome to another episode of Manufacturing Matters, the occasional video series where we at TECH B2B look at key factors that are affecting the manufacturing technology sector. As always, I'm here with my good friend and associate Jimmy Carroll over TECH B2B and today we're lucky enough to have Ronald Mueller with us from Vision Markets.

Now, Ronald, for those of you who don't know is a key player in the machine vision industry, he's also going to be presenting a keynote presentation at the Vision Systems Design Tech Summit, which is scheduled for November 4th, which is a virtual event. And during that presentation, he's going to be delivering his keynote From Illumination to the Cloud, New Technological Answers to Unprecedented Customer Needs. And for us, we thought when we first heard about it, this is really an exciting thing to share because as you know, in our last episode, we were looking at artificial intelligence and deep learning and its impact on  manufacturing and kind of real worlding, trying to get some ground truth. What are the key pieces that still need to be in place to make deep learning an effective solution for manufacturing? So today we think that Ron's presentation has even more insights to that looking at edge processes versus the cloud and the whole vertical stack that needs to be in place. So with that, I'm going to ask Ron if he would mind just summarizing a little bit for us his upcoming keynote presentation.

Ronald Müller, Vision Markets:

Yeah. So as it happens, this presentation could have been inspired by my recent visit of the Vision Show and what I've seen there. And I think most of the people who were there can confirm that there is in our industry, still a large variety of suppliers providing a large variety of solutions, hardware solutions, software solutions and in a summary, this all are very valuable and amazing bits and pieces of the solution that an end customer in industrial automation is needing. And in the presentation, we will go a little bit more about what the end customers are looking at, what they are looking for. And we will relate that to what is existing currently in the market and how can we as a community so to say as players in this market, how can they be developing solutions that are then attractive for the customers?

Jimmy Carroll, TECH B2B Marketing:

Thanks, Ron. So, from talking to you, it seems that your presentation is based on some recent projects in the automation industry, specifically on Industry 4.0 strategy and how that can provide a fully integrated solution for major manufacturing companies. So we're talking cloud services, interconnectivity, deep learning, machine vision, and so on. Can you talk a little bit about these technologies and how they work together and why this topic is so relevant today?

Müller: Yeah, sure. Yeah, that's correct. So what we have used to say those conclusions that we show in the presentation is, as you said, in several projects with industrial automation manufacturers, if I may say. Now, these players are seeing the needs of the customer, they see the big picture, right? They see how can industries and can factories be more flexible? How can they be more effective? How can they be replicated? All right. So we have this situation that it's not enough today to have just one single factory somewhere in Southeast Asia and rely on the fact that they are shipping goods at a decent rate. So that's not working. We have seen in the in the very recent past that this is not reliable. So big OEMs and end users want to replicate their production lines have one in Asia, one in U.S., one in Europe, in the best case to maintain their capability to deliver. And that's creating new challenges, right? So that's creating new needs on the customer side and they are needing now, especially with the advent of deep learning, there could benefit from cloud-based services where you have central instances, how you manage images, how you manage models, how you manage annotation data, how you deploy those types of models across one factory or multiple factories. And that's not just the case for deep learning. There are plenty of other use cases for cloud services. Just think about documentation purposes for quality audit, so to say. And also the rollout and management of the edge devices, which is something that is just so to say, happening for big solutions and provided by big players today. But more and more even smaller manufacturers are demanding that.

Hardin: Ron, real quick, we know a lot of companies who are exploring cloud services have concerns about cloud security and putting their production data up into the cloud. So can you speak a little bit about that? I know you're going into more detail in your presentation, but what can you share with us today?

Müller: Yeah we are still living in different worlds, right? So there is one type of customer, it's not even necessarily regional differences it's more like a culture thing of many companies, some say we don't want our production IT at all being connected to the internet or if so, just one or three times a week for, you know, making some some major updates. So for this type of customer, they are limiting so to say their benefits that they could gain from cloud connectivity. At the same time, it's still feasible to and there are new technologies now to establish security gateways, a multi-layered gateways between the production IT and the internet. And also, obviously, you can host a lot of information, a lot of managed let's say cloud services on premises. You have seen maybe some announcement with collaborations between IBM, Red Hat and so on, who are now able to replicate on premises cloud services, etc. So there are new technologies and services evolving. But again, it's a zoo of different players, some playing only in the cloud, some providing only software, some providing only hardware. And for the end customers, it's an enormous challenge to get together all these different pieces and finally get the solution that they want. And it's enormous effort for integration and enormous effort of discussion, supplier selection, and what have you. And actually, we it's a good idea I'm sure that we can relieve the end users from those pains. And in the presentation, I might give some outlook, let's say, and some suggestions and inspiration how we can help them.

Carroll: Ron, I look at the interconnectivity between products or technologies used in Industry 4.0 application, where all devices are connected. In some cases, or in most cases, large manufacturers will already have a lot of these technologies in place, but they may not have them all connected together. So I kind of think of this in the way that I think about a deep learning system where sometimes people just don't know where to get started. So if a company has a lot of these technologies already deployed, how do they get them to work together in the way that you're talking about? Because in a lot of ways, it's Industry 4.0 is more process-based than it is product-based.

Müller: Well, with enough money and budget and effort and time, you can make things work, right? So we have deployments of deep learning solutions in big plans and there are ways to get that to work. The thing is that the market and the penetration of the market with such types of services could be significantly higher if we made it easier for even big customers, but also smaller customers to get something to work. And we are in the machine vision world, we are quite well connected. Everybody knows that the key players of the competition and of other, let's say, vendors. So there is a strong network within our community. What is missing is the part where we are connecting to big players. As I said, like the IBM's, the Red Hat, the hyperscalers who can help us and actually who we will be needing to some extent for providing a fully stacked solution. And that's something I think where initiatives can go to. That's something where networks and partnerships can be established within our industry. And then we can be approaching hyperscalers. And I mean, you know how it is at the end of the day,it's very hard to aim high at the very beginning and you don't know where you are ending up. So one approach to the whole story is certainly to pick a lighthouse project with a lighthouse customer, determine the needs and requirements that they have and then pull together a network of partners that are so synergetic. And by that, we can kind of create such a framework that is applied at one instance, and out of that, the solution can then be evolving towards a larger ecosystem, larger framework where more people are feeding in. And at the end of the day, this is what the end customers are actually looking for.

Hardin: Well, I'm looking forward to your presentation, sounds like there's going to be some call to actions for the OEM and the vendors within our market space, as well as some design best practices and business organizational best practices for approaching these types of complex projects that require or will benefit from some cloud integration. So thank you so much for sharing your time with us today, Ron. We really, really appreciate it.

Müller: Been a pleasure. Thank you.

Hardin: We look forward to having you back on some future episode. Thanks, everybody. See you soon.