“The whole dialogue has shifted from what can it do to when it will and how do we scale it?” 

Humanoid robotics technology is slowly but surely moving into real-world deployment. The excitement behind humanoids is just one of several hot topics in robotics discussed on this special episode of the Manufacturing Matters podcast. 

In this episode, TECH B2B Marketing’s Jimmy Carroll sits down with a powerhouse panel of robotics and technology leaders to unpack what’s actually happening in the industry right now — and what’s coming next. Joe Gemma, CEO of Integrion Automation; Dr. Daniel Lau, professor and director of graduate studies in engineering at the University of Kentucky; and Robert Little, chief of robotics strategy at Novanta Inc., share candid insights on investments into U.S. manufacturing and the impact this will have on robotics and automation technologies as companies continue to navigate the labor shortage. They also discuss physical AI, humanoid safety and potential applications, digital twins and simulations, mergers and acquisitions, and how automation technology overall continues to evolve and advance.  

Jimmy Carroll: [00:00:00] Hi, everybody. My name is Jimmy Carroll. I’m the vice president of operations at Tech B2B Marketing. And welcome to this special edition of the Manufacturing Matters podcast, where we’ve compiled an excellent, and I say excellent, I’m personally very excited to talk to everyone here, our panel, to discuss robotics and the state of robotics. And if I could introduce everybody to my left. Next to me, I have Dr. Daniel Lau, Bob Little, and Joe Gemma. Guys, if we could start with Dan and just introduce yourself, what you do. You’re a newly anointed IEEE fellow, so congratulations.

Dr. Daniel Lau: [00:01:03] Thank you very much for that. Yes, I am a professor and the director of graduate studies for electrical engineering at the University of Kentucky. Yes, 25 years and my area is research in 3D imaging and accurate color reproduction, along those lines.

Jimmy Carroll: [00:01:17] Robert?

Robert Little: [00:01:19] Yes, 40 years in the business. I co-founded ATI Industrial Automation and led it. Today I work for Novanta, who purchased ATI four-and-a-half years ago. And our expertise is robotic force sensing, robotic tool changing. And our sister company, robotic servo drives and actuators and encoders.

Joe Gemma: [00:01:44] Hi. Thank you for inviting me. I’m excited to be here. My name is Joe Gemma, and I’m very fortunate to be the leader of the Integrion Automation business. As the CEO, I lead a wonderful staff of people, and we are an integrator in this space across multiple markets and trying to help people through their automation journey. That’s what our company does. Thank you again for having us.

Jimmy Carroll: [00:02:06] Absolutely. It’s my pleasure entirely. So, I want to ask, I’ll just put the question out there and whoever wants to pick it up can answer. And then you guys, everyone else can chime in as you need or as you want. I’m curious to know what are some of the most notable developments in robotics over the past couple of years, in your opinion? And whoever wants to take that one first can go.

Joe Gemma: [00:02:25] I’ll jump in to get it started. I think one of the biggest leap has gone from what we have called predictive AI. And that’s kind of like guessing when something’s going to break and trying to, you know, anticipation of that. Where now it’s going to what we’re calling agenetic AI, which is more really managing the workflow and actually fixing it in the workflow process. I think that’s a big change in the last couple of years that has impacted manufacturing in a very positive way.

Dr. Daniel Lau: [00:02:57] I myself am on the robots is the vision systems. You know, when I first started teaching at the University of Kentucky, my first time of flight image sensor was $50,000, and it was 320 by 240. And today you get 1024 by 1024 for under two grand, fully IP67, and the like. And you know, the factory floor is certainly where you expect to see that type of AMR. But just like in my home, robot lawnmowers and the robot vacuum cleaners now have what, when I started my career, very sophisticated vision systems that I never would have imagined could be in such a small form factor. Yeah.

Joe Gemma: [00:03:40] Great point.

Robert Little: [00:03:40] Yeah. And let me just add on to Joe. The last two years has been a massive investment in physical AI and humanoids. Just the other day, Skild AI out of Pittsburgh is building a foundation model that will go to all robots, that will be physical AI. And they just received $1.4 billion.

Joe Gemma: [00:04:06] And to add a little bit to what Dan said, I think one of the other changes is, I call it edge AI intelligence. So, it’s kind of gone from the cloud to more localized and simply people will recognize it from a perspective of like the sensors where you can go into a work cell, but the robot can react because these sensors are becoming the eyes and the ears of that in milliseconds. So, you can really make changes on the fly tat we weren’t even able to do. So, these things that we talk about in our home lives actually have come out of the manufacturing environment and made that process much quicker and adaptable.

Jimmy Carroll: [00:04:44] Yeah, absolutely. You guys jumped in right away, and you talked about AI and humanoids, which I love, and we’re going to loop back to that a little bit. But if we continue to get a little bit granular, what are some other notable developments in technologies like end-of-arm tooling, AI enablement, mobility, things like that. Anything in particular that excites you or you’ve seen that’s innovative that you know it’s worth mentioning?

Joe Gemma: [00:05:09] One thing, again, if you guys don’t mind, I’ll jump in, something called digital twin. So digital twin has been around for a little while, but it was more on the fringes of utilization. But that technology has really been more embraced across all types of manufacturing where it can really utilize that, that technology. And basically what it’s doing is it can animate a manufacturing process, so you can accelerate the development process. So, if they have a new part, they want to see if their system can run that part or what cycles time. You can do it basically in real time. It’s emulating the exact manufacturing process. So, it’s being utilized across so many different areas today that not so long ago it was seen as almost a black art.

Robert Little: [00:05:52] And let me add, safety has become such an important topic, something Joe knows very well. At Novanta, we just completed a servo drive specifically with safety protocol on it for humanoids that is going to be a bit of a game changer. We also finished a force torque sensor specifically for humanoids. So, the humanoid market as, you could say controversial as it is, is really driving a lot of component developers. I mean, the money that’s going into end-of-arm tooling today is at the highest I’ve seen it.

Joe Gemma: [00:06:38] Agreed. You know he talked about end-of-arm tooling today that have haptic capabilities too. So, it’s you know, Soft Robotics is a good example. They’ve really evolved, and it’s software driven, for sure. But that capability of fragile, it’s helped a lot of companies that look at the high-mix, low-volume processes. Because of that technology where these types of end-of-arm tooling could really become the eyes and the ears of the robot, like we would with our hands to our own human aspect of it.

Joe Gemma: [00:07:08] Another thing I think to mention, you know, you’ve probably heard me say in the past, I’m not a big fan of the term cobots because it’s really collaborative applications. But that technology and more from a macro view is really seen a higher acceptance because people are starting to recognize it’s not a replacement technology; it’s an augmentation technology. And because it recognizes, it can work alongside humans to give that process a little bit better creativity to the worker. But the mundane part to that and utilization of that in so many different areas has been adopted because now people accepted it not taking jobs. And that’s been a challenge we’ve had in our industry. And they recognize it helps the job instead of taking jobs.

Dr. Daniel Lau: [00:07:51] Now something real quickly, something I’m really excited about now is it doesn’t make it into the automation floor is just the use of AI models to accelerate finite element analysis and other types of physics-based simulations. Because some of the AI architectures are just on a, you know, 15 to 30 times faster than running the traditional physics calculations. So, I’ve been doing a lot of work in that area myself.

Jimmy Carroll: [00:08:20] It’s always really interesting to me to see all these different technologies working together to solve new problems and something that comes to light, that came to light for me recently, was I went to a Schneider Electric event, and they had these sessions where they were talking about innovations in plumbing and water cooling to fuel these data centers that are used for things like digital simulation. And I’m like, that’s not something I would have thought of in the past. But even in here, you look at different, you know, you see the different-colored lanyards and you talked about how AI adds flexibility and vision adds new elements for robots, and all these things work together to solve different problems. And it’s kind of just to me, that’s always been a really fun thing to follow and see how that changes over time. One thing that Bob, I have to admit, like following you on LinkedIn. That’s where I find a lot of my my news the quickest. There have been some real notable investments in robotics lately. Where do you think robotics, and, Bob, I guess we could start with you on this one, where do you think robotic investments are heading? And what’s that mean for U.S. manufacturing?

Robert Little: [00:09:20] Well, the money’s going into physical AI right now and safety. I should mention safety is absolutely critical. And it’s not about humanoids. Believe it or not, humanoids is a subsection of physical AI. You know what really got me thinking is when I saw FANUC come out with a physical AI statement and a partnership with NVIDIA. They’ve been a partner with NVIDIA for 12 years, but they just committed to creation of physical AI applications and they opened up the ROS 2. Who would have thought FANUC would have ROS 2 on their platform? Open with a one-millisecond latency. Unbelievable. And that is a game changer. But not just for FANUC. I’m talking about all the robot companies are deeply involved, you know, and but the biggest one probably this year that we need to make a point of — ABB being purchased by SoftBank specifically to develop the next autonomous robotics.

Joe Gemma: [00:10:33] Yeah, I think Bob’s made a really important point that a lot of what we’ve seen in the investments and mergers and acquisitions has been kind of a full-stack strategy. I mean, even large companies like Rockwell, Teradyne investing in those technologies have that full-stack in house because the software, the hardware, etc., has now become a necessary process for those companies to own and help be able to be successful in the markets. So, I think a lot of the merger acquisitions, just like the ABB and with SoftBank. Softbank has been building the acquisition strategies for multiple companies to be able to support what they’ll be able to bring to the market. I think that’s a real trend that we’re going to continue to see, too.

Dr. Daniel Lau: [00:11:13] You know, I was going to say on the academic side, I have the ability to look even further out than some of these guys. And I do want to say that humanoid robotics has me very excited. I mean, when [00:11:28] you look at the idea of 24/7 work, and what I’m talking about isn’t in the factory floor, but if we start talking about like dairy farms, something I’m very interested in, you know, the humanoid robot, it’s just so fascinating. If you can do certain jobs with a humanoid robot, it just opens a whole world of possibilities for other industries and agriculture in particular. [00:11:56]

Joe Gemma: [00:11:56] I’m glad you brought that up, because humanoids, [00:11:59] I think that that whole dialogue has shifted from can you do it to more when it will and how do we scale it? And if you think about it, it’s almost zero-position retrofitting. You know, there’s aisles in these warehouses because they’re built for humans or a humanoid can transcend that same space. It can go in an elevator. You know, that technology. And interestingly enough, along that you see that some of the leaders in using that technology are some of the automotive companies. BMW, Mercedes have invested heavily. If you think about the automation industry over the last 50 years, the leaders almost always in adopting those technologies has been the automotive industry.  [00:12:40]So, that’s to me, it’s a testimony of that will be the next thing. Maybe it’s not tomorrow, but it is in the near future that there will continue to be investment in those areas because the automotive industry sees that these things can be where humans are. Again. I like the term zero retrofitting. You can utilize the same space you’re in with the humanoid, just like a human being.

Jimmy Carroll: [00:13:01] You know, you guys talk a lot about safety though. With safety with humanoids. How do we get to the point where humanoids are considered safe, and you can use them on the factory floor or in the warehouse or even in the home?

Robert Little: [00:13:14] Well, I would say there are two types of safety. There’s one where we want humanoids to work without fences but not around people. And so the way that’s being resolved is through special sensors, which is a combination of lidar and vision, which is up your alley, and you probably know very well. But that gets you only so far. The point is, is that when a human gets too close to a humanoid, humanoid goes into a safe mode. Okay, so the humanoids can work without the fences, but it can’t work with people. The Holy Grail is that work with people. But I’m going to let that is your answer.

Dr. Daniel Lau: [00:13:52] Oh. That’s not. I am not going to take responsibility for making sure.

Joe Gemma: [00:13:58] So, we’re here at the A3 Business Forum, which is the association. Actually we have a safety component of this industry, the A3. And they are working on standards for humanoids because it is a valid point. Safety standards. It is a valid point. We have to recognize what that looks like, and safety certifications strategy around that risk assessments. A3 has been a leader in that space, and the traditional automation space in manufacturing, we will be a leader in that same space for humanoids because it is a valid point. We want to understand that we keep people safe. So, let’s drive what that standard should look like. And we are working on that currently, so that we can have those standards that people can follow and keep us safe, both in the manufacturing environment and at home.

Dr. Daniel Lau: [00:14:44] And you know, eye safety for Vision Systems has been because that has been drilled at such a high responsibility for laser safety, and yeah, I’m certain safety will be one of the number one issues that will determine when people and humanoid robots can interact.

Jimmy Carroll: [00:15:05] Sure. Yeah. There’s a number of topics that were brought up there that I want to ask questions that kind of go different ways. But one of them I want to circle back. I want to talk about lights out and 24/7 like we brought up. But in terms of some of these investments in robotics, how do you think that bodes for, like US reshoring efforts? It’s going to, obviously, automation has to play the strongest role in reshoring, right? So, how do you think that’s going to impact US sales of robots in the next couple of years?

Joe Gemma: [00:15:34] Well, I think it certainly, Bob’s going to talk to us because he’s an expert in that area. I think one thing that really is unique today, because of automation, we have bridged that wage gap. Today because of automation, we can build a high-tech manufacturing facility in South Carolina and compete with the Southeast Asia impact because of logistics and shipping and be in a very competitive space, because of utilizing this technology of automation in a high-tech manufacturing space. I will say one thing that companies that are investing that have to recognize, it’s still important to have that critical level of ability in your staff. In fact, Erik Nieves, who you know well, he uses the term person-in-the-loop. Wwe still need to have a person-in-the-loop. So lights out, so to speak. I’m not completely sold on that. I think it’s more of a person-in-the-loop to make sure that process works, but skilled. So, it’s creating another level of skilled technicians that we will need, and it will actually create jobs.

Robert Little: [00:16:35] Good point. You know, let me just say something to the audience here. We are in the middle of a manufacturing revolution. You may not know it, but we are in a massive upscale for manufacturing. In my opinion, very similar to the 1939 preparation for World War II. We have allowed ourselves to economically be dependent on other nations for goods that we may not be able to be dependent on in the future. The time is now to start. And it has started with $600 billion going into pharmaceutical. There must be a new pharmaceutical facility announced every two weeks, and each pharmaceutical facility starts at $500 million to $3 billion. That will ensure that we can get our medicines locally. Semiconductor. We’ve built some fabs. Intel has built a fabulous Fab 52 that makes a new chip that’s excellent, and they’re looking to build a new one. But then take TSMC out of Taiwan, built their first fab in Arizona. They got ten more coming. Those fabs are like $50 billion minimal apiece. Texas Instruments. They spent $50 billion on fabs just recently. And they’re spending more. The bottom line is this: we’re going to see a massive amount of reshoring, a massive amount of rebuilding where new lines are coming in for automotive, for appliances, for semiconductors, for pharmaceuticals, for mining, for agriculture. It’s happening. We absolutely need robotics. We’re 500,000 people short today. We’re going to be 3 million labor shortage in manufacturing in the next five-plus years because of people like me retiring. I’m the problem. But the solution — automation and robotics.

Joe Gemma: [00:18:42] I agree, it’s obviously, there’s statistics. There’ll be a net gain of millions of jobs in some of those, even this industry. In fact, I saw the term robot orchestrators. It’s going to be high-tech, high-skilled jobs. So, it’s not just bringing jobs and utilizing this technology. It’s going to be high-skill, high-paying jobs. So, that’s even more an advantage of what’s going on with the reshoring aspect in America.

Jimmy Carroll: [00:19:08] You said you’re not completely sold on on lights out. And I like that idea because, you know, human-in-the-loop, person-in-the-loop. It’s not just for AI data training, right? It’s also for all these other things too. Why aren’t you, why aren’t you sold? And what will it take for us to get there? Not only on the factory floor but in warehousing too. Like lights-out warehousing? How far off are we? What would be needed for us to get there and to scale it?

Joe Gemma: [00:19:31] Great question. I’m not sure we’ll ever get to what we would call, people would like to see lights out, and lights out, meaning no people at all. And there’s just too many variabilities in the world that affect things. You know, a truck backs up to the dock wrong. It’s just not aligned properly. So then the equipment that was supposed to be able to unload that truck can’t. So, you need a person-in-the-loop to say, well, you know, we need to realign this or we need to reposition. But it might be as simple as talking to the robot system and say, shift 30 inches to the left type thing. And I do think that’s going to happen, but I think that’s why we need, we’ll always need somebody, you know, person-in-the-loop. Things happen, things break. You don’t anticipate, it happens. You know, we have robustness of processes and products but still things change. So, I think we’ll always need some person-in-the-loop. I do think we’ll get less and less of the needed, I’ll say the more of that blue collar aspect of some of these jobs. There will be more of a high white collar type aspect to it. I don’t think it’ll ever be completely lights out.

Jimmy Carroll: [00:20:38] Mostly lights out.

Joe Gemma: [00:20:39] Right. And it’s funny because when Bob and I were early in our career, that was an expectation of automation. Was lights-out manufacturing, and it actually, the software of today didn’t exist. So, it had a lot of promise that it couldn’t live up to it. And it really challenged people to even maybe, you know, potentially lose their jobs because they were trying to design around lights out, which really couldn’t exist. Software has gotten better and better, but to the ability that you would need complete lights out, I think is, can never say never, but it’s certainly a challenge.

Robert Little: [00:21:11] And let me say, we need more people in manufacturing today than we ever, ever have. And if you’re looking for a new career, I highly ask that you consider manufacturing and that there is a ton of jobs in manufacturing from automation and robotics, as we’ve talked about, to quality, to planning, purchasing, assembling, and hordes of highly skilled technical jobs like welding.

Joe Gemma: [00:21:40] And Bob makes a good point. If you think about the volume of things that we as consumers use today. As a manufacturing aspect, it’s just more than quadrupled in terms of, it’s exponential. And I’ve said before, and you’ve probably even heard me say it, you know, we used to have a phone mounted to the wall, and everybody used the same one, and if you wanted it to be a little more mobile, you put a longer cord. Today we all carry one phone or two. We have multiple televisions in the house, and that’s true of appliances, etc. Those have multiplied. You got the home vacuum cleaner that Dr. Dan talked about. So, there’s all these things that the sheer volume of what we use as consumers, it really means that manufacturing capability. But the personnel is more than manufacturing.

Jimmy Carroll: [00:22:24] Dan, as a professor, you get the chance to work with a lot of young people, and I don’t know how many of them go into manufacturing, but maybe some of them are electrical engineers that may become integrators. What are those conversations . . . How do those go with people when you talk about, you know, pursuing careers in manufacturing?

Dr. Daniel Lau: [00:22:42] Well, you know that being in Kentucky, the most of the vast majority of our students are interested in power producing. How are you going to produce electricity? What are the upgrades going to be necessary to make all these facilities possible? So, that’s driving a lot of what’s happening at the University of Kentucky. But without question, there’s growing . . . We have faculty. We don’t have a robotics engineering program standalone, but we in electrical engineering have robotics faculty. Mechanical engineering has robotics faculty. And then there are faculty in CS that are also focused on robotics. And so AI as well. You cannot get a graduate degree without a serious AI component to your education. The undergraduates, not so much. You’re getting subject matter expertise as undergraduates. And when you go to graduate schools, you get the AI component. So, students, I think are very, very exciting. But yeah, you know, the research aspect of my job is just shifted so much into the direction of AI.

Jimmy Carroll: [00:23:45] Yeah, I’m not surprised at all.

Dr. Daniel Lau: [00:23:46] Yeah, you just, you know, you just can’t get away with doing, like, structure. Like, there has to be an AI component of what we’re doing now. Yeah.

Jimmy Carroll: [00:23:54] Fun question for you guys is, what predictions do you have for the next couple of years? And, and I know I spoke earlier to Stu Shepherd, and he said, well, you know, anything that’s going to come out in the next one or two years is already being worked on, but in the next ten years . . . So, maybe it’s one or two or maybe ten, but do you have any strong convictions when it comes to things you expect to see in robotics in the next, however many years?

Joe Gemma: [00:24:18] I think I agree with Stu that it’s already started. I think we’re at the beginning of what I call a standardization war. You know, you got things like Tesla and humanoids. We talked about these different types of equipment on the manufacturing floor that the consumer of this technology, they want things to talk to each other. So, we’re going to have to kind of like Bluetooth capability or Wi-Fi capability where, you know, the early days that was, you know, you didn’t have everything connected to each other.Today, more and more in manufacturing, they want things to be able to talk to each other. So, I think that called standardization war. I think we’re at the cusp of that. And coming out of that, we will have some system just like our phones. No matter what phone you have, you can talk to, it doesn’t matter the brand, you can talk to each other. I think that’s what we’ll see in the next couple of years. I think that’s going to evolve. I don’t think it’s ten years. I think it’s the next couple of years.

Robert Little: [00:25:12] And, you know, I’ve talked about the manufacturing renaissance. We’re going to see a lot more reshoring and rebuilding, which is going to boost robot sales. You’re going to see a lot more of that. But the second thing that’s going to really help robot sales in the next one or two years, you’re going to start seeing some of the physical AI applications coming out into the market. Keep your eyes open. We’re going to see some really nice stuff starting to come out. It’s going to be less talk and more action.

Dr. Daniel Lau: [00:25:42] And I was going to say you’re going to continue to see the autonomous self-driving vehicles. The cost of lidar, which doesn’t really fit in, but the cost of lidar has plummeted so much that I actually think that even though I think it is possible to have camera-based self-driving cars, the cost of lidar is just so low I don’t know why every car isn’t going to have lidar. It’s just getting so cheap. But yeah, that’s the only thing I would say on the sensor side. Yeah.

Joe Gemma: [00:26:12] And I’ll add one last thing too. I think this generative AI aspect will drive more and more of a really quick adoption changes in the manufacturing space, where you’ll get an opportunity, maybe not in the next couple years, but in near future, where an engineer can say, reprioritize the blue product and package that for the next three hours and just tell the system. And it’s getting to that point very soon. I don’t think it’s ten years out. It may not be in the next one or, you know, 24 months, but 36 to 48 wouldn’t be a big surprise.

Jimmy Carroll: [00:26:45] Selfishly, I’d like to keep you guys here for another hour or two, but I want to be respectful of your time. And we’ve got a first-timers meeting that’s closing out in a few minutes. Anything that we haven’t talked about that you feel like you want to bring up or you’re excited about?

Robert Little: [00:27:01] Well, I’m curious to see, once the SoftBank purchase of ABB closes, how that forms and changes ABB. That’s a curiosity thing. But there’s so much new stuff coming out, I’m just excited to learn about it and and talk about it.

Joe Gemma: [00:27:21] I think I mentioned earlier, I think we’re going to see more of this merger acquisition strategy as well, and building that full-stack, end-to-end capabilities in these companies. So, I would keep an eye out for that. I think you’ll see more of that in the very near future.

Dr. Daniel Lau: [00:27:37] And I just, you know, I’d like to see more and more students applying to engineering education.

Jimmy Carroll: [00:27:43] Absolutely.

Robert Little: [00:27:44] And I second that.

Jimmy Carroll: [00:27:46] Yeah, absolutely. For sure. Well, if anybody has any questions, I would recommend reaching out to these gentlemen on LinkedIn. Or I would be happy to pass them along. Guys, thank you so much for taking the time. I really appreciate it. If you’d like to send any questions or comments, or you want to be on the podcast, or tell me I did a bad job, then visit us at manufacturing-matters.com and otherwise, thanks so much for tuning in.