Frustrated manufacturing software vendors await someone or something to lead them. If not Elon Musk, then perhaps AI.
The conversation addressed a wide range of questions. To wit: Are the benefits of Industry 4.0 right around the corner, if we can only figure out how to get there? Would vendors be better off lowering their sights and focusing on smaller, simpler needs within the manufacturing space? Can manufacturing software make big advances if the manufacturing hardware could keep up? And how does AI fit into all this?
Engineering.com received exclusive access to the roundtable transcript, which appears below in edited form.
Hosted by Ron Fritz, CEO of Tech Soft 3D, who also served as moderator, the roundtable participants were:
- Sam Burgess, Founder & CEO at SamsonVT, which provides a 3D platform for aftermarket parts and components, as well as interactive work instructions.
- Andy Cheadle, Chief Technical Officer at CloudNC, a precision parts manufacturer and developer of CAM automation software.
- Milan Kocic, developer of the open innovation platform Sixth Sense from digital solutions company Hexagon.
- Suchit Jain, Vice President of Strategy and Business Development at SOLIDWORKS.
- Tyler Barnes, President at Tech Soft 3D, a provider of software-development tools.
Ron Fritz: It’s the year 2023. When it comes to the high-precision manufacturing area, what makes you want to tear your hair out when you think about problems that still haven’t been solved?
Andy Cheadle: Everyone talks about “Industry 4.0” and digital connectivity and the like—and yet factories and shop floors often simply aren’t ready for true digital transformation. These terms have been buzzwords for a while, but the fact is that many customers I visit have not evolved/been adequately enabled to a point where they can deploy automation manufacturing technologies successfully, so we can find it difficult to bring customers along to a more advanced way of doing things.
If I want to go and digitally enable a factory, I’ve probably got to go and grapple with its ERP [enterprise resource planning] system, MRP [material requirements planning] system, PLM [product lifecycle management] system and its design tools. The message here is that as companies build up this big estate of very capable heterogeneous but disconnected software, it can actually hold them back a bit as far as embracing new and transformative ways of manufacturing.
Milan Kocic: I feel like a lot of people innovating the future have never actually been to a factory. For example, people have been talking about MBD [model-based definition] and PMI [product manufacturing information] for years now. In my world, metrology and inspection are super important—but we did some research and found that 75 percent of factories still use a 2D printout with a highlighter to inspect stuff. So, it seems like a lot of innovations are being prepared for customers who aren’t actually ready to receive those innovations, as Andy alluded to.
Suchit Jain: While you were speaking, Milan, it got me thinking about Elon Musk and Tesla. Elon had the idea to create a new kind of car that was software-driven from the ground up. You can’t retroactively put those software innovations into any old car—the hardware has to keep up with the software.
In the same way, maybe someone like Elon Musk needs to enter the high-precision manufacturing world and create something like a CNC machine that’s been designed from the ground up to be software driven. If it’s still the same old machine, it won’t be able to keep up with other advances on the software front.
Sam Burgess: I’ve seen similar things in my world, which is the world of aftermarket parts. When we first approached big customers, we actually thought that maybe they wouldn’t need our software—that surely, they already had something in-house that created a digital parts catalog. But no, with the exception of maybe a single factory, they’re still using paper-based catalogs for aftermarket parts.
Tyler Barnes: At the end of the day, it’s inherently a conservative group that we’re delivering solutions to. On the software front, think how long it took customers to cross the chasm from 2D to 3D design—nearly 15 years. It just takes time for adoption to happen for any new innovation.
Ron Fritz: It seems like some of what we’re circling around here is that while Industry 4.0 is getting all the attention as “the next big thing,” there are some very basic needs in the manufacturing world that currently aren’t being met.
Milan Kocic: I think the next big thing is probably a lot of little things. I think a lot of corporates spend time talking about the next big thing but not taking care of the little things. I think if you do innovation in those little concepts, eventually you solve the bigger problems.
Sam Burgess: There’s 2.5 times more profitability in parts sales than in new equipment sales, but hardly anybody focuses on the efficient manufacture of aftermarket parts because it’s not the shiny, exciting new toy. I can count the number of times that we’ve spoken to big, large OEMs where we’ve asked for just a simple SBOM [service bill of materials] for components. You can hear the screams in the background where people are running around to make one. They don’t have the information. It’s not a priority for anyone to automate this. This is a very basic manufacturing area that needs to be addressed.
Andy Cheadle: When it comes to automating manufacturing to unlock and drive efficiency gains, my thoughts are that if we are not opinionated about how we manufacture a part choosing quality/time trade-off, etc.—then ultimately, we have to build our future CAM software packages the traditional way that gives the human in the loop every little fine-grain bit of control.
But that’s not how you achieve automation! It’s by being opinionated, making some decisions, removing some elements. Keep in mind, it’s not about labor elimination, which people worry about when it comes to automation. It’s about enabling the human to do less of the repetitive, low-value tasks and actually free them up to do the higher-value tasks.
Suchit Jain: There are some companies like Proto Labs, whose bread and butter is manufacturing parts, who have managed to achieve a fairly good level of automation and moved away from 2D drawings. If every company who produced parts were like them, that’d be great—but they’re more the exception than the rule.
Ron Fritz: MBD and PMI were mentioned earlier as providing a level of information that is critical for automation. All the CAD systems today are pretty darn good at generating PMI, but there are a couple different barriers we’ve touched on that are getting in the way. It might be a technological barrier, like the fact that the machines aren’t sophisticated enough to run the software that can interpret the PMI, or it might be a cultural barrier like “we’ve always done things with paper—that’s how we do things.” Are there other things that are stopping PMI from driving automation?
Milan Kocic: I think part of it is ubiquity—or rather, lack thereof. I think less than 20 percent of companies have PMI for their 3D data, so people wind up reverting back to 2D printouts.
Suchit Jain: In the construction industry, for example, there is the issue of accountability—and paper trails still seem to be a preferred method of having a single version of truth, particularly in a fragmented environment where not everyone is working for the same company. I think that sort of mentality is quite pervasive everywhere.
Andy Cheadle: That fragmentation that you refer to. It starts because people have historically said, “I’m a designer and then I’m going to pass off to the manufacturer.” And in doing that, they create these barriers in the process. Now, we are beginning to see people say, “Let’s put the designer and the manufacturer together and allow them to iterate on the design and carry that conversation across and forward.”
Ron Fritz: Let’s look ahead five years from now. What areas do we think we might say, “Well, that’s sorted now”? There’s a lot happening in AI right now, for example, which could change things.
Tyler Barnes: I think that AI probably will change things in small places all along the design to manufacture process, with software that helps you make better decisions, but isn’t making all the decisions for you. That goes to the concept of generative design and being able to work through tons of different design concepts, through to analysis. So, I think that as a technology, AI probably has the biggest potential to impact the greatest number of different things along the entire chain.
Sam Burgess: Honestly, people no longer using WhatsApp for 90 percent of parts transactions would be a huge win, as far as I’m concerned. On another note, equipment downtime can be a real headache for manufacturers, resulting in lost production. But the good news is that AI is a powerful tool that can help prevent these issues and reduce costs. I believe that by leveraging AI-powered algorithms to monitor and analyze equipment control and sensor data in real time, manufacturers can quickly identify and diagnose upcoming failures to proactively address them and prevent downtime altogether. At SamsonVT, we’re working on an AI solution that can help solve these problems and revolutionize the manufacturing industry. By reducing downtime due to equipment failure and cutting unnecessary maintenance time, our AI-powered solution can help manufacturers maximize efficiency, minimize costs and keep their production lines running smoothly. I am super excited to announce that as soon as we’re ready, so watch this space!
Andy Cheadle: I probably will have failed in my job if I have not brought an appropriate level of “single-click to make” or some degree of autonomy around CAM program simplification and automatic part generations. We are getting there, though. So, I’m going to put my money into that.
Suchit Jain: I’m very bullish on AI in general for a lot of different fields. In terms of automation of manufacturing, I think that within five years, AI will easily be able to help humans make better decisions around things like costing by saying option X is going to be more costly, or option Y is going to be less costly.
I don’t know that we’ll see a big game changer on the machine/hardware side of things—that Elon Musk scenario I was describing earlier, where someone just creates an entirely next generation of machine—but I think we will see a lot on the software side and it’ll be a bunch of small things, as Milan was talking about earlier.
Milan Kocic: At the end of the day, improving manufacturing is all about reducing the choices people have to make. By that I mean: If you get people 80 percent to their goal and then let them make decisions for the last 20 percent, that addresses a lot of issues.
I agree with everybody that we’re at an inflection point where AI—with enough training—can get people to that 80 percent threshold. There’s enough data coming in that you can train it relatively easily to do that.
What I don’t think we should focus on is trying to get to 100 percent and totally replacing humans—that’s not achievable in our lifetimes. However, focusing on smaller things will affect a great deal of what happens in the future. I think we’ll see a lot of interesting developments within five years.
