MFG Austin Ep. 11 Nathan Byman, CEO of Wolfram Manufacturing

Nathan Byman is the founder of Wolfram Manufacturing, a company based in Austin, Texas that is not your typical machine shop. While they do make an array of fabricated and machined parts for customers, through their deployment of technology and advanced techniques, they have also developed a consulting practice around automation and deep tech. Nathan shares his insights about the intersection of old school manufacturing and new innovations that are driving change in subtractive manufacturing. He holds and engineering degree from the University of Texas and an MBA from Rice. Wolfram Manufacturing was founded in 2011 and employs about 20 people.

That’s a great question. Right in the beginning there was not actually a drive to start the company. I very much wanted to find a company to buy. It was a little bit like business school training. You know, you do a leverage purchase, you improve it a little bit, you grow it from there. That was the direction. But I was working in manufacturing environments, and the companies I was with were buying and bringing together other manufacturing companies and I was on the end that was helping to level up their technology. And the whole time I’m thinking, this will be great for when I go out and I find my small little company that I can buy.

Well, the more I was implementing those technologies, the more I was seeing how hard it was to take a company with a lot of inertia and change it over. At some point, a switch flipped, and I thought, you know what? I want to start with the core set of technologies and just business principles about how we hire people and the types of people, and I want to build it from scratch and I think I’ll get there faster.

Now, at the time, I also thought that I had a great line on several years-worth of business. And, you know, it was a done deal right up until the point where our first machine “which was— I mean, we went big. We went in with a $600,000-plus machine. It was about to land. Down payments paid, financings lined up, and all that work evaporated.

It was an oil and gas customer. The oil and gas customer was like, “Wow, we’ve been planning for this for the last two years, this buildup of this new product. But it turns out that we accidentally entered our forecast three times and you know, we went back and audited it and we realized the demand’s not there after all.” So it just disappeared. And that was rough.

Oh man, that was a hustle that we had not signed on for. And it was, I’m not going to lie, it made Wolfram a bad word around the house for years. You know, I love technology. I love talking to people about technology. I’m an introvert at heart. So getting out there and selling was not my game, not my business partner’s game. But it was what we needed to preserve our investment. So yeah, knocking on a lot of doors, pulling on friendships and relationships, trying to show that the technology we had put together would work.

Now, we had one machine that we started with. It could output as much as three machines, not because it could machine faster, but the set of technologies we had put together really multiplied what it could do. But that was a huge impediment. Talking to people, they were like—you have one machine. Well, we know one machine is only good for this many parts. And convincing people of that was tough to start.

Oh, man. Well, let’s see. Today is May 10th, so…. I say that only half tongue in cheek. You know, as an entrepreneur in business, you’re never exactly sure. But you know, we were in that just really painful hustle for probably five years. And then we were able to hook up in oil and gas at the time, and we would have a good six months, a bad six months, a good six months, a bad six months. It was like we were an aircraft trying to get off the runway. And it’s like, okay, we’re up. Oh, you know, the wind just pushed us back down and we bumped again. And okay, we’re still running for the end of the runway, but are we going to make it over the trees at the end? I would say it is the last few years here that we really felt like we cleared the trees and now we’re up seeing some scenery in deciding where we want to go next.

I feel like a lot of entrepreneur entrepreneurs might be this way. I don’t tend to look back and reevaluate things. I feel like if I went back and ran numbers, it would’ve made a lot of sense to have a book of business. It would’ve been much less heartache. But I don’t think we would’ve built what we have today.

What we have today feels like so much opportunity. We have a way of manufacturing that I personally feel like is the future. We have companies like Lockheed Martin and folks coming in and looking at what we’re doing and they’re saying, “Oh man, we make incredible stuff, but we need to be here with our manufacturing.”

So, we have created something and sustained something that is good. It is good for manufacturing and the more we can share it and help people… people are seeing it and they want it. And that’s great. I mean, it will be rewarding for us, but there’s also a reward in helping manufacturing in this country get better.

And then we have kind of this product stack. We make stuff, then we can go out and help people do things better, all the way up through automation. Like this low level automation, robots. We can help people get there with clarity and with them seeing that there is a path and they can see it in our own shop, and I think that’s awesome.

And then, one step beyond that, we can talk to and represent this whole product stack. Now we are launching our own software that ties it all together. It lets you see your whole manufacturing operation. Not like ERP, but actually what is the heartbeat of your shop. You know, there’s a whole communications infrastructure to it. It’s like we’re teaching the machine so that when they have a problem, they reach out and let us know that they have a problem so somebody can go work on it.

So, we have thousandth of a horsepower, horsepower sensors on all the machines. If a tool starts to consume more power than it did historically, it will send a signal to a controller that we’ve installed on the machine. That signal will then be sent over to a software that we are running that will reroute that to the appropriate Slack or Teams channel that represents that machine.

We have display boards on the wall and everything else. But the machine will almost literally ,just like one of our coworkers, send us a message that says, “Hey, this tool is wearing a little bit early. It’s only made this many out of this many parts. Somebody should check on it before something goes wrong.”

Now, one step farther, if something does go wrong, it will stop before any human on their best day ever could. It will stop faster than that. And it will stop without damaging anything, in most cases. And then we get a more urgent message that says, “Hey, this machine just decided to stop to protect itself.”

And it gives us all the details about it that redirects our actions. It lets us keep people moving efficiently.

You know, it’s a mix. It is all available to people. We use a lot of technology from Caron Engineering. A lot of the sensors and the feedback systems are theirs, where it has some personality and it’s being monitored and communicating back out. That’s product that we have written and kind of perfected in our own shop. But that’s also available, so we sell that and we can help other people with it too.

It’s kind of fascinating. You have some of the first cool, highly automated machines in machining and manufacturing. But in some ways it’s like the industry that time forgot. Like G-code that we use to drive the machines is from like the 1950s and people are just continuing to do it almost exactly the same way, to make all this other super cool technology.

You know, we’ve got iPhones and just incredible technology packed into things, but the way we make all this stuff might be on equipment from the 1950s. It’s fascinating and that is one of the big things that we are kind of working to change. So, in addition to machining, like the other things that our business offers, are really around bringing that change around.

You know, the core machines, aside from additive, which is a little different, but the core machines of machining are going to keep doing their business. That’s part of why they’re still around. They do it well. But, some of the other markets we serve and what is changing in automation, there are machine tending robots and that’s good. That’s a helping hand. But if you really want to free people from the process and supercharge the process, then what we’re looking at is like deep automation. I mean, going in and having the machine, while it is cutting, know what’s going on and make better decisions to keep the process going. When you run manufacturing fast, things break. So people will run things fast and they’ll keep people around to try and stop things in a pinch. When you automate things, most people slow things way down. So what we do is, we put in this really low level automation where we are monitoring vibrations of machines, horsepower, force. We’re monitoring everything and we learn that ,we train it, and we will have the machines adjust to keep those, basically keep cutting conditions optimal. And slow down, speed up, stop the machine if something happens. So this really low level automation, it basically, it lets us run everything faster than would be possible any other way. It protects our machines, it protects our parts, and it sets you up to do robotic automation and be successful. A lot of automation with robots just fails. People don’t realize how much they’re touching their process. And if you still have to have a person getting into a robot cage to touch it, then you’ve defeated having a robot at all.

Yeah. When we started the shop, we were doing it all for our own benefit. I had seen, and practiced, and worked with, a lot of these different technologies and decided to bring them together. We started our company and we started using the technologies. And the more people would come to our shop, see what we were doing, they’d ask us if we would help them.

So now, really from the absolute mom-and-pop shops to the biggest names—just absolute titans of industry—we help make their manufacturing better by bringing in this really deep automation. So on that front, the business is aerospace, satellites, rockets, cars, defense. We get to work with and, help some really cool companies on that front.

You know, the last few years have really seen a pickup in the ease of use of classic industrial robots and cobots just taking the market by storm. It’s still not trivial. But there is just tremendous application for tending robots. And they are just helping out all over. If you’re running a machine that is dropping out a lot of parts in a hurry, we have them picking up the parts, running them through inspection a couple different ways, and setting them aside. We have things with short cycle times that have kind of an exotic loading process. Instead of tying a person to a machine, particularly with the manpower challenges of the last couple years, it is just super valuable to let a robot basically keep us from tying a person to a machine where they have to touch it every 10 minutes, to loading a stack and coming back six or eight hours later. We can use that manpower somewhere else. And those things have been just super successful. And along the lines of the manpower shortages, it helps you in a couple different ways. Everybody is struggling. So one, you have to find people. But then, everybody has their own unique kind of personal supply chain challenges right now. Like they get sick with COVID themselves, somebody in their household gets sick, they have to take care of someone. It’s fascinating how many things are pulling people away from their work life. So, automation can help just iron that out. But also, as we’ve seen pay rates go up, and the housing market take off, and everything else that is making everything just so expensive. I mean, record gas prices again today. You know, the more automation you bring in, the more we get to be competitive with the salaries that we have. So, we are still seeing growth, we are still hiring additional people. But the more people we have, with robots as a kind of a force multiplier, the more successful everybody is.

Oh, you know, the world of inspection still has a lot of room to grow. There are some kind of awesome technologies coming out now, but anybody in manufacturing knows that inspecting and ensuring that you have good parts is one of the toughest things to do. But you have also some super cool technology coming to bear. That’s probably one of the biggest growth markets—just an interesting inspection technology that can let you automate those things even better.

That’s a really good question. You know, if I’m going to stick to some of the things that I know the best—if I’m around metal machining and the automation of that—one of the most important things to do is understand how stable their process is to begin with.

So I will encourage people if you can stand and watch your part, if you can watch your machine run and nobody has to touch it. Like, as a shop owner or a manufacturing engineer, just sit in a chair and watch your machine for a day. And if a guy doesn’t have to tap something, or adjust something, or tighten something, then maybe you’ll be okay. But make sure that your process is settled first. As we’re just trying to make this world automated, those are two of the things that we focus on: that deep level automation of sensors and feedback, and software to see if machines are able to do their jobs without interruption. Once you know that, you can try and get to cobots and robots to start getting the parts in and out.

And I’m not going to try and make it sound too rosy. There’s a huge proportion of people that will buy things like these cobots that are “easy to program.” And they’ll buy them. They’ll start trying to implement it and they will give up because it is a slightly different language. It turns out programming the robot to move around is super easy. That is truth in advertising. It’s super easy. Make it move, make it grab.

What is still a real challenge is how it interfaces to whatever machine it’s interfacing to. So, who is the lead dance partner who gives a signal that the other one takes their cue from? How does the other one properly say that I am completely ready for the robot to come in? There are just any number of challenges there. But that world is also changing significantly. There are two companies that are working on really cool technologies on that front. One lets you talk to any different robot so you can have a mix of robots and have this easy programming interface.

And then there’s another company that is basically putting together a similar software for machine tools that lets you talk to any machine tool and see where it’s at in its process and send the right signals. And that last connection between those two is historically the hardest part of a robot implementation. Either working with some of those technologies that I just mentioned, or, hiring a good integrator to do that last mile, who actually has familiarity with your machine tool and the type of robot you’ve picked, is pretty important.

It is generally excitement. Our perspective is a little different because we are running a shop, have run a shop continuously, and are running a shop. So when we go talk to people, it’s almost like people have a trade detector or something. When you speak and you have pain in your heart about something you’re describing to them, they see it come through. And just that mutual understanding normally gives us a good starting point. And then when we talk about things and we can be enthusiastic about it, it helps people get over that skepticism. We normally connect pretty well.

Just as a general rule, love additive manufacturing. And all its various faces have a place everywhere. Even if you are a shop that is doing all metal subtractive machining, it absolutely pays for us to have even the cheapest, plastic filament printing available. We have lots of printers around and there are printed parts everywhere. So it is just a force multiplier.

We teach everybody in our shop how to design in CAD. We keep inexpensive CAD on shape available for everybody. We teach them how to how to use it. We teach them how to slice and print even if it’s their own hobby stuff. So that when they are out in the shop looking at how to make our processes better, they start to think, not, “I wonder if I could use this tool to make this job easier.” You flip it so they start thinking, “I could design the perfect tool to make that job just so much easier.” We just have printed odds and ends everywhere.

Beyond that, when you get to metal additive, that one is interesting. When it first came out, I thought it would mature a lot faster than it has. The applications that can benefit from it have not been as wide scale as everybody had hoped. Having the perfect thing printed that takes a little bit longer just because of the physics of laying down and heating and cooling and everything, have not been able to compete except in kind of your edge design cases. So hopefully we’ll continue to see that change.

Something neat— I saw copper has historically been very hard to print additively. I’ve seen some chatter going on about some new alloys that have come out in copper that are printable. And things like that will start to change what we can do for getting heat out of things. And this is pretty low-level, but things like lithium ion batteries, we’re getting energy density, which is great. Brushless motors we’re able to make stuff happen. Heat rejection is still a big problem. So, as we can do like exotic copper printed things, I think we’re going to get better at heat rejection, and we’re going to see better hand tools, better machine tools, better all kinds of things. Five years ago, I would’ve bet that we would’ve been completely disrupted by metal machining, metal additive. But it turns out that machining is remarkably resilient. A lot of stuff is still going to be done the same way for a while.

Oh man, if we have our way, then machining is going to be kind of cool and fun again.

Make machining cool again. Yeah, that’s what I’m going to put out there. There are all kinds of YouTube channels and influencers who make things, you know. And I feel like people are connecting with that and seeing that. But then when you translate that to big production, no two ways around it, kind of the heart and soul, it’s not exactly the same. Right? You know? So how do we make that part fun again? And I think, you step back from the machine, you add some of these elements of automation to it. You have people looking at whole processes instead of just machines, how they can improve things, giving them better tools. I like to call it making it like video game machining. So, you know, you have all this data access available. You can see your tool health meter, and how they’re doing. It’s all tracked. And let people win at it, making things better. And I think it becomes more fun for everybody. I think we’re finally at the point where we can level up and change what machining and manufacturing is, and I think we’re going to make it cool.