Speaker 0 00:00:00 Thanks to the explosion of collaborative robots. Robots are now working alongside humans, getting them to work intelligently with each other is the next big challenge.
Speaker 1 00:00:11 Hello everyone. And welcome to the robot industry podcast. My name is Jim Baretta, and I'm your host. And I'm really thrilled today to have Neil Cardella, who is president of energy technologies, a division of universal robots, which is a Teradyne company. Energy focuses on developing software for co operative robot motion planning and simulation energies. Software platform. Acton is used by government agencies and commercial companies to simplify development and speed deployment of multi-arm high access systems in the medical, industrial energy and space robotics sectors as a co-founder of energy, Neil helped navigate and bootstrap the company through various phases from having an initial focus on government R and D developing software for NASA and the department of defense to a commercial engineering services company, then to a profitable software licensing services company that was ultimately purchased by Teradyne in 2018. Neil holds a BS in electrical engineering from the university of Hartford and an Ms in computer science from NYU Tandon school of energy, which is formally Polytechnic Institute of New York university. Welcome to the webinar,
Speaker 0 00:01:26 Neil. Thanks, Jim. Thanks for having me.
Speaker 1 00:01:30 Um, one of the, we, we kind of chatted a little bit before about some of the questions that I wanted to ask you. And, um, I wanna maybe have you tell us a little bit about energy and the services from kind of a high level?
Speaker 0 00:01:43 Sure. So energy is we've a relatively long history, uh, energy. Got it. Start developing software for NASA, um, almost 20 years ago. And the focus of the company back in the early days was to create a software framework that could be used to, um, coordinate, um, many axes of high degree of freedom robots. First project that we worked on was a, um, for a robot called robot that was being developed jointly by DARPA and NASA, and that robot had 57 degrees of freedom. So it was a very complex robot. And when you start to get robots that have that many joints, you have to start thinking about how do you choose how to move the joints to have the robot do something that you want it to do? It's a real challenge. So NASA wanted a solution to that, but they wanted a solution that could be a platform that was generalized enough to be used for other robots.
Speaker 0 00:02:43 And out of that, um, early work in the government, R and D sort of phase of things, we developed our product acting. Um, first half of the company's, um, history really was focused on these government R and D projects starting with, uh, NASA, uh, first, not NASA JSC, then JPL, uh, then, um, DARPA, the department of defense and, uh, about, um, I'd say now CI's about eight, 10 years ago. We started to see demand for this type of technology in the commercial sector. And as is often the case, the government is almost like an investor of a last resort for new technologies. That might be a bit ahead of their time, but we started to see, uh, companies that were looking to do things in robotics that, uh, they couldn't do, and they didn't have software to do it. And we transitioned the company to focus more on, um, on commercial robotics. So we provide, uh, software and services for this area of robotics that we call cooperative robotics, which is robots that have, um, you know, a high number of degrees of freedom and are maybe multiple robots working together. And that's what we were doing when we were acquired by, uh, Teradyne in 2018, we were actually doing the project for them and decided to, um, we had some discussions around the potential of an acquisition after they saw what the software could do.
Speaker 1 00:04:15 Very nice and firm. And from your perspective, what's happening in the kind of robot, uh, maybe Cobra, HUD motion control industry right now?
Speaker 0 00:04:25 Well, you know, it's, uh, industrial robots have been around for a really long time. She's it's over 50 years probably that there have been industrial robots and historically robot manufacturers controlled both the hardware and the software of a solution. So, um, you know, you look traditionally at an industrial robot manufacturer, you know, take your pick KUKA Denzo ABB. They all sort of had their own software stack and their own hardware stack. Um, we originally focused on and we still do focus on startups and companies that were developing a new robot and needed software for those new robots. But what we have seen, there's been a shift probably within the last five, six years. A lot of it driven by Ross and Ross industrial for some of the industrial players to open up, um, to the possibility of third-party software, um, supporting their, their hardware. So that's a shift that we've seen. And, um, we obviously were part of universal robots and we apply our software to universal robots, robots, but our software is, um, robot agnostic and it can be applied to other industrial robots and other customer months. So that's a shift that we're seeing. I love that
Speaker 1 00:05:45 The term robotic Gnostic, uh, it's, it's a very exciting time in the robot industry. And we talked a little bit about cooperative robots. And what did you mean by that?
Speaker 0 00:05:54 Yeah. So if you think about it universal robots, they really, uh, they pioneered collaborative robots and collaborative robots are robots that are designed to work with humans. And in the way that we're used the term cooperative robots, cooperative robots are really robots that are designed to work with other robots in an intelligent manner. And you might say, well, robots, I look at an assembly line and I see lots of robots working near each other, but that's a little different than them working with each other. And we believe that a, a significant portion of tasks are underserved by automation simply because the robots that are attempting to automate these tasks are too low in terms of degrees of freedom. Uh, humans have two arms. And, um, if you try to do any tasks right now, without that, you know, without those two arms, you'll see that it's a very difficult thing to do. So, you know, simply stated cooperative robots are robots working with robots or robots working with other, um, other automation equipment.
Speaker 1 00:07:07 And how does having more access, help solve industrial automation problems?
Speaker 0 00:07:15 We think that the there's really an under appreciation for the value of degrees of freedom. If you think about us humans, we take for granted the fact that we have 200 plus joints evolution has designed us in such a way that we have, uh, significantly more degrees of freedom than your traditional industrial robot, which is usually six degrees of freedom or less six axes or joints or less. And, you know, if you think about it, there's a parallel here with where things have gone in the AI and deep learning industry where, you know, perceptrons were, you know, the whole concept behind, um, uh, artificial neural nets that came out in the sixties. And it was modeling a bit of the human brain in terms of axons and dendrites. And, um, it kind of went out of Vogue for a while and then it's come back.
Speaker 0 00:08:06 You know, people realized, you know, evolution had something right here. And the fact that we have the amount of joints that we do says that it's more than just cognitive ability that is required to perform tasks. And, you know, if you want to think about the power of degrees of freedom, like I mentioned before, uh, think about doing any task that you do as a human, but do it with one hand tied behind your back. And then furthermore, put your one, um, let's say useful arm in a cast because you have seven degrees of freedom and most robots don't. So you're even limited there. And you'll notice that it's very, very difficult. It's not a cognition component. You could have the smartest robot in the world, but if it is six degrees of freedom, it's going be extremely limited in what it will be able to do.
Speaker 0 00:09:01 Now, one analogy that we like to use is if you've ever seen those, uh, arcade crane games, where you've got a, you're playing a game where you're trying to pick a prize and you've got a joystick and you can move it in a couple of different directions and have it go up and down and open and close the crane, it's kind of a tricky game. It doesn't work. It's not so easy. Um, but you're a human brain and you're manipulating a very, very simple four degree of freedom robot. And even though you have a human brain, you're very limited in your ability to be successful at that grade at that game. So, you know, we, we, it's more than cognition and we think that there's a bridging the gap is enabling higher degree of freedom applications.
Speaker 1 00:09:49 No, it's a great example. Um, so getting back onto the commercial side Neil up, who are kind of your typical customers, you mentioned startups, but at the beginning we also mentioned medical.
Speaker 0 00:10:01 Yeah. So if you think about the concept of cooperative robots, it's robots that are working with each other. And when you think about, um, surgical and medical robotic systems, if you can see pictures of like the DaVinci system or other systems, there are very frequently many robots that are in close coordination with other robots. And this is typical. And that's an example of an area that can leverage and use our software. Um, other areas, uh, are say the energy sector, we've done a lot of work in the energy sector on, um, uh, supporting a company that was building a fully autonomous oil and gas drilling rig using software where to perform, uh, drill string assembly operations, where you're doing something called like a trip in, or you're pulling out a drill string, a trip out, it requires the coordination of many different robots working together. And so in these types of applications, this is an area that we, uh, we see our software offering unique value. Okay.
Speaker 1 00:11:04 And so one of the problems that you solve for your customers are obviously time to market,
Speaker 0 00:11:12 Right? So, you know, we, we really focus on, um, developing the algorithms and software that a customer can use really focusing on the robotics portion and allowing our customers to focus on the application, like build the intellectual property in the area that they sort of have the most, um, expertise in, and they can assume away the complexities of the software. So we really, uh, think that that's a strong, um, value proposition. What we're offering is getting somebody to market faster because they don't need to build something from scratch. Uh, they have a proven solution that's tested and a proven partner in energy to help get to market more quickly.
Speaker 1 00:12:02 And so how do you work with them? You, it, you, is it a statement of work or are you using, um, uh, are you using Ross in some cases?
Speaker 0 00:12:11 Uh, we, we can, we typically don't use Ross, but we certainly have, um, we have sort of ways that we can integrate with Ross. And we do have customers that use Ross. Um, our engagements typically are, um, almost like, you know, we consider ourselves a partner and we can support a robotics project from the full sort of product life cycle, starting with, um, preliminary design and analysis of the new robotic system all the way through deployment. So we provide software development kits and software tools and also engineering services. So that we're kind of supporting this customer in this area of robotics, path planning, motion control, cooperative, robotics, um, taking that piece of what they're trying to do, working with their team and helping them get to market more quickly with a robust solution.
Speaker 1 00:13:05 And how about, how long does it take for a PR? I know, I know this is probably a hard question, right? Is there like an average deployment? Is it a year? Like, I, I come from, uh, uh, complex, uh, robot integration systems and it's always, everything seems to take about a year. How long are you in this?
Speaker 0 00:13:21 Yeah, I mean, it, it it's, since we often work with OEMs that are building a new robot or say a new surgical robot a year is, you know, reasonable to short, in some cases, it really is going to depend. Um, but these are generally long engagements. Um, and depending on when we start with the customer, that's going to depend on how, uh, you know, how quickly things move forward. Um, and it also is going to depend on the industry. For instance, if it's a surgical robot, there are certifications five, 10 , FDA approvals and things like that have to be considered. So those can take a long time on the flip side. Um, we have gotten customers up and running with a, let's say a standard universal robot. That's trying to do something, uh, in this cooperative space in a matter of, uh, you know, a few months, which is, uh, you know, if you, if you weigh that against the complexity of the type of solution, it's actually a very short period of time. Okay.
Speaker 1 00:14:25 Thanks, Neil. So I want to talk to you a little bit about a specific problem in the industry and it has to do with real-time path planning. So for our audience out there, can you explain what is a real-time path planning and why is it so hard?
Speaker 0 00:14:38 Sure. Well, uh, most industrial robots today are programmed to do one task, uh, over and over again, the same way every time, and usually create that path. You'll use offline programming tools that gets downloaded into the controller, and it works just fine for many applications where you don't need that sort of changing of the, um, the path of the robot real-time path planning, um, is it really focuses on the modification of the robust trajectory on a time step by timestamp basis based on sensor feedback. So, uh, think of a robot that has to do something a little bit differently based on four sensors, like force feedback coming in, or a vision sensor, maybe moving out of the way of another robot as an example, or picking a part off of a moving conveyor, whenever you have to alter their trajectory of the robot at runtime, that's what we refer to as real-time path planning.
Speaker 0 00:15:47 And there's lots of practical implications of why you need that with cooperative robots. So I'll end, our customers are using some of those. So for instance, and the oil and gas drilling example, I've given you earlier, um, being able to let's say, put a, um, a drilling a drill bit on the end of a drill string and torque that in, well, that requires a careful analysis of the torques as you're screwing that in. And the trajectory of the robot needs to be able to sort of accommodate that, uh, for dual alarm handoffs, you're usually limited in the accuracy of the robot. Even if you have perfect calibration, you may have small differences in as-built components of the parts. You have errors in the calibration, so that you're going to need feedback. It could be forced feedback, it could be vision feedback. It could be a fusion of both. So these are the types of reasons that you need this real-time path planning. And this is only an area that's going to grow in the future. As we start moving more toward, um, let's say less rigid, more semi-structured environments.
Speaker 1 00:17:03 And of course it must get kind of complicated when you're underwater or you're up in space, or you've got gravity working against
Speaker 0 00:17:12 Absolutely. And that's why you need, you know, just like humans. We don't do we know if you look at a human, a human does not have accurate and coders. If I close my eyes and move my arm around, let's say I close my eyes and try and pick up this glass of water on my desk. I'm probably not going to be able to do it all that. Well. I'm probably going to miss a little bit and I'm going to end up compensating based on some other sensors, not the encoders of where I think the joints of my arm are. And if you think about how a human works, when you pick something up, you're doing a visual surveying loop, uh, you're looking at your hand and what you're trying to pick up. And every, maybe 30 frames per second, you're doing slight corrections of the trajectory in order to do that. So, you know, right now, industrial robots, they're reliant on super accurate encoders and super rigid hardware, but really what that is, is a crutch, because we don't have the ability to perform this real-time path planning and accommodate for alternate sensory feedback, um, in an effective way in most of these industrial robot applications.
Speaker 1 00:18:22 Nice. So I, one of the questions I kind of, it's always interested me and fascinated me is, is a little bit more about medical robots and, and surgical robots. How, how mature is the industry and are you seeing a real uptick in your business in that sector?
Speaker 0 00:18:38 Uh, so to answer your question on the maturity of the, of the business it's, um, if you think about intuitive surgical, they launched way back. I think, you know, before 2000, 1999, and they did a great job in paving the way for surgical robots, um, and also they dominated the market. You know, they're kind of the universal robots of surgical robots, I would say. Um, but they also showed the value of surgical robots, which was super important, um, around the 2016 timeframe, a lot of the original patents that intuitive surgical put in started to expire, and that sort of spawned a bit of a race to building a competing system. And there's been a lot of large companies and startups that have been sort of moving into the surgical robotics space in the last, uh, four or five years. And we have seen an uptick in, um, and interest in this area from, you know, our customers. We're, we're getting certainly a lot more, uh, interest from startups and established companies that are trying to bring a medical device to market a medical robot to market. And again, they want to focus on the, the very unique aspects of the, um, robots that they're building versus the motion control components.
Speaker 1 00:20:08 Sure. And, and, and of course having a trusted partner is, is a very valuable thing to them. So where are these medical robots used? I'm assuming it's always in north America. Is it in Europe? Are they used in any other specific environments?
Speaker 0 00:20:24 Um, so I would say yes and yes, there are definitely in north America. There's quite a few in north America. There's quite a few in Europe. Um, there's, uh, a lot of medical, uh, uh, robotics work going on in Europe, um, and a lot in the United States. Um, it's also a lot in China. You know, China actually has a pretty robust, uh, medical robotics space as well. So, you know, we're really seeing a lot and there, um, if you're asking kind of where, you know, where they're being used, you know, there is a lot of, uh, new areas that, uh, robotics is starting to work within the medical, um, medical device field. I would say some that we've seen are, um, my microscopy where, you know, you might have just a, you know, a robotic system to support, uh, imaging imaging is actually a very big area right now. Sure. Um, you know, others like laparoscopic surgery is, is, uh, still very big. That's probably the biggest, there are, there are type seals, which is a single insertion type of surgery, which is a little bit different, has a different set of problems. And orthopedics is another area that we're seeing a big uptake.
Speaker 1 00:21:42 That's great. Thank you. So this, my next question is kind of more about how you innovate your own work and how you work with your partners.
Speaker 0 00:21:53 Well, you know, the innovate piece is a good one because now that we're part of a large company, we are especially sensitive to innovation. And how do you do that? A, um, it's a big challenge and it's something that we've, uh, we've always prided ourselves in being a very innovative company. We were innovative before we were acquired. And we're innovated now that we are part of a larger organization. Um, what we really try to do is, you know, first and foremost, we, we really hire the best and the brightest and we work to empower them, um, to do what they do best. So within the energy organization, we're still highly collaborative. We, um, we hire specialists that, um, and we allow them to test out and try new innovations and new ideas. And we do that within a framework. And this is really, really important that allows for us to sort of present that innovation in a robust fashion to our customers.
Speaker 0 00:22:59 Um, what we really try to do is, um, make sure that, you know, we're working in lots of areas that require, um, software that is very, very robust, very well tested. And sometimes you will see there is a balance between innovation and robustness, and it's a, uh, it's an important aspect of, uh, running an innovative organization to strike that balance in the appropriate way. There's no shortage of robotics, startups that failed with lots and lots of very, very smart people. Um, because they're, they're actually exceedingly good algorithms and technology wasn't really ready for mainstream and never got to mainstream. And there are, um, innovators that, that don't necessarily want to take a new algorithm and take it to where it needs to be from a product worthiness perspective. And so that, that's kind of the challenge that we've, I think we've kind of figured out how to do that in the right way. And all of our engineers are, um, cognizant of our end goal with this product. And the result is that we've delivered a product that's innovative and robust, which I think is unique in this space.
Speaker 1 00:24:25 Oh, that's great. Thank you for that. Our sponsor for this episode is Earhart automation systems, Earhart builds and commissions turnkey solutions for their worldwide clients. With over 80 years of precision manufacturing, they understand the complex world of robotics, automated manufacturing, and project management, delivering world-class custom automation on time and on budget contact one of their sales engineers to see what Earhart can build for you and their
[email protected] Earhart. It's hard to spell E H R H a R D T. I'd also like to thank and acknowledge our partner, a three, the association for advancing automation. They're the leading trade association in the world for robotics, vision and imaging motion control and motors, and the industrial artificial intelligence technologies visit automate.org to learn more. And I'd like to thank painted robot painted robot builds and integrates digital solutions. They're a web development firm that offers SEO, digital and social marketing, and can set up and connect CRM and other ERP tools to unify your marketing sales and operations. And there are painted robot.com. And if you'd like to get in touch with us at the robot industry podcast, you can find me Jim Beretta on LinkedIn. We'll see you next time. Thanks for listening. Be safe out there. Today's podcast was produced by customer attraction, industrial marketing, and I'd like to thank my nephew, Chris gray for the music, Chris Colvin for audio production, my partner, Janet, and our partners, a three painted robots and our sponsor Earhart automation systems.
