Creative ElectronGET A DEMO

X-ray News

Fireside Chat with the Xperts: Robots and Cobots – the Good, the Bad, and the Ugly

Robots and Cobots – the Good, the Bad, and the Ugly is the subject of this Fireside Chat with the Xperts.  It’s no Spaghetti Western, but Griffin Lemaster does provide a terrific look at industrial robots, with a brief history and an introduction to their role in modern manufacturing.

From defining the difference between a cobot and a robot, to discussing the many ways in which both compliment human capabilities in manufacturing, this chat packs a lot into a brief 30 minutes.  Once you’ve enjoyed this one, you can check out the full line-up of upcoming as well as archived Fireside Chats here.

Transcript:

David Kruidhof:

Good morning, everyone. Welcome to another Fireside Chat with the Xperts. We are here today with Griffin Lemaster, our wonderful COO. He’s going to be talking to us today about robots and cobots, and how they can be used to augment your X-ray inspection, supplement your workforce, enjoy a safer environment, things like that. So we have a really good presentation here. I’m really excited to hear what Griffin has to say. Take it away.

Griffin Lemaster:

Thanks, Dave, for that warm introduction. As Dave said, my name’s Griffin Lemaster. I’m the COO here at Creative Electron.

Griffin Lemaster:

All right, perfect. So today I’m going to be talking about robots and cobots, the good, the bad and the ugly. So just kind of a quick rundown, going to kind of go over a little bit of the history of robots, where they came from, why we’re using them, get into the differences between the different types, discuss the differences between robots and cobots and everything else that’s out there, and also then hit on the how we use them in our industry in X-ray inspection and what we’re currently doing with them, and then we’re going to touch on some of the advances that are coming out here in the future. Some really cool stuff. So without further ado, let’s jump right in.

Griffin Lemaster:

In the beginning, right? So robots have really been a science fiction concept since really the earliest records of humans, going from the ancient Egyptians and the ancient Chinese that they all left their depictions of automata and other inventions that resembled human and animal likeness. So the myth of these creations really has transcended the generations, right? And produced itself in all forms of entertainment that we see. So you see the picture on the left, that’s Talos. He’s the towering bronze statue that protected the island of Crete from invasion from enemies. And then on the right, you see one of the more famous robots from cinematographic history, the robot from the movie Metropolis in 1927. Really cool special effects they did back then for not even having computers, so if you haven’t seen it, it’s worth a watch.

Griffin Lemaster:

But for the sake of understanding this fireside chat and us all being on the same page, we’re going to get away from the whole sci-fi concept of robots and humanoids, and we’re going to stick with a very simple definition, and that is going to be a programmable device used to perform one or many tasks autonomously. And this is really just to… I don’t want to convolute the theory of what we’re talking about with simple motorized mechanisms or remote control devices. These are not simple machines. These are programmed and to do a certain set of tasks or a certain set of movements without consistent human intervention and human interaction to control these, and so they’re operating on their own.

Griffin Lemaster:

So why robot? Why have robots become so popular? Why are we using them as much as we do today? It’s been an industry demand, so different industries, automotive, you see the image there, I mean, that’s kind of what we think of when we think of industrial robots, is this automotive robot going together and putting cars together on an assembly line. That’s a really good explanation of why we are where we’re at with them. It’s dangerous working conditions. It’s repetitive motions, repetitive tasks. It’s very high speeds and very highly repeatable and precise movements. And these are tasks that we as humans aren’t suited for. We can’t do these by ourselves, so essentially robots, these industrial automation robots, were created to fill this capability, this lack of capabilities that we have with humans, and to add to it.

Griffin Lemaster:

So kind of getting back a little bit to the history, where industrial automation started and where it came from. So back in 1937 is really the first modern programmable robot, and that was built from toy parts. The inventor of the Meccano toy sets created this first robot. Meccano was like Erector sets, but just a little earlier. But what he did was he built essentially a crane-like structure that had six different degrees of freedom running off one motor, and it was programmed by this paper with hole punches, which really became synonymous for programming for a long time afterwards, and so that kind of really paved the way for a lot of future programming, a lot of future of robots.

Griffin Lemaster:

In 1960, there was the first digitally programmable robot sold to GM, and so the purpose of that was to essentially lift really heavy metal castings, hot metal castings, and stack them. And so not a complicated process, but something that really operators couldn’t do on their own, and it’s automated a process that didn’t need to have intervention. And so we see this entrance, this really start of industrial automation, come through the automotive field. So in 1996, first cobot was invented, and so that was… We’ll get into a little bit more about what cobots are and what that means.

Griffin Lemaster:

And then 1998 to the present, we’re in this current robotics boom. And so over the last 20 years, we’ve seen essentially a 600%increase in annual production, so just massive amounts of robots that are out there doing industrial automation. It’s really driven by the automotive market, but really close behind automotive is electronics manufacturing, and so this boom is really being driven by a couple key select markets led by automotive and electronics, and for different reasons in different tasks.

Griffin Lemaster:

So these applications within these markets. So automotive, it’s doing some rough and dirty work, right? You’re doing welding. It’s doing grinding. It’s doing lifting castings, working in harsh environments. But what we see in the electronics industry is really doing a little more of the finesse of the electronics market. So pick and place machines, if you’re not familiar, is the machines essentially put every single component that’s on a circuit board, populate that circuit board before it gets soldered. And so it’s doing very fast, very precise movements over and over and over, and so it’s a very high-finesse application, and then very quick. And then essentially after the circuit board gets soldered, you have this quality inspection process, automated optical inspection or automated X-ray inspection, AXI, where you’re doing a quality inspection after the fact. Cameras moving around, taking images, looking at the solder joints, looking at orientation and alignment, things like that, and basically making a pass/fail criteria on that, and then doing whatever it has to, pushing that down the line to the next process. And then also simple like product handling, right? It’s lifting. It’s twisting. It’s inspecting, loading and unloading.

Griffin Lemaster:

So that’s kind of a big difference between automotive and electronics, and then medical is this whole other beast. Really there’s a lot of new advances coming out of this robotics push in the medical field. So from diagnosis, essentially using CT scan, that’s a diagnostic robot, right? And treatments and even medical device manufacturing.

Griffin Lemaster:

So let’s get into this second definition, the cobot. The cobot really is, it’s a collaborative robot designed to work with humans in a shared workspace. So it is a robot, but it’s equipped with a unique set of capabilities that the robots really don’t have, and almost a limitation at the same time. So now that we know the definitions, you guys are… you’re all set, and we’re good to go, right? Of course not. Probably brings up more questions than you had before. How do you know which to use, cobot or robot? Is it just color preference? What if I don’t like green or yellow? Is there other choices, right? So let’s highlight some of the key characteristics of the differences between these robots and cobots.

Griffin Lemaster:

So let’s start with the cobot. I call this one the good. Because cobots, they’re designed to work in the same workplace as humans, their motion is highly regulated. So they don’t move as fast. They have sensors built in to sense any type of small collisions or interference. And yes, the colors on the robots are there for aesthetics. They’re to make it seem a little more appealing and less intimidating. Same with the corners, they’re all rounded and curved. Even the material that they’re built from is a little more lightweight, because it experiences less forces and less acceleration than a standard robot. And these are so the robots are your characteristic… They’re bound by the classic traits of being a good guy, right? They are no frills, but they’re gentle and they love their mother. I mean, it’s a safe choice if you will.

Griffin Lemaster:

But because there’s this option for human presence and interaction, there’s also a lot less restrictions and operational requirements, so you can walk right up and deliver a tray of parts to a cobot, and you can even pat it on the shoulder if you want, and the cobot’ll load the parts, pick up a part, orient it the correct a way, load it into the system, and the system will take it and do whatever the next step is. Because they interact with humans so regularly, you can do that. If it was a robot, however, there’s a lot more restrictions on essentially what you can do with a robot in a tight space. So, I mean, for smaller companies, possibly that don’t have as much area to work with or they don’t have as many resources to work with, cobots are actually a pretty good option to get into this robotic automation, because it doesn’t require a set of extra conveyor belts to load parts. It doesn’t require all these interlocking fences around the robot in order to work in a safe environment.

Griffin Lemaster:

So let’s go to the bad, the bad boy robots, right? Like I just said, these are your bad boys. These are the fast-flying, do-or-dying, drop them off and leave them crying. They are tools for Industrial automation with a capital I. Robots, they’re there for the grunt work. They do your heavy lifting. They do grinding, welding, extreme speed work. And these are all things that we need in these rugged, tough production demands. They can operate at the bottom of the ocean. They can operate in an 800-degree environment. And so they’re just built tough, right? But one of these limitations working with robots is it really does have to be protected from human presence. You have to build interlocking fences so that you can’t really get hit by one of these things, because if it malfunctions in any way, shape, or form, or usually it’s human error that goes and bypasses an interlock or gets in the way, these robots have enough power and enough speed to do some serious damage or even kill a person.

Griffin Lemaster:

So talked about the good cobots and the bad boy robots, and both of these are standalone devices used very commonly in industrial automation. But under the skin of even more automation platforms out there, you typically find these in-house creations. That’s why I call them the ugly. They’re not as polished. They’re not as clean. They’re not as refined. But these in-house creations that make it do a limited number of tasks, right? You build them for your task that you need them to do, but they usually can do these really well. And I’m speaking about every widget maker, everything thingamabob winder that all uses arrays of electric motors, pneumatic cylinders, conveyor belts, PLCs and too many sensors to count. So these are ugly creations that really make the masterpieces that we build, really kind of the backbone of industrial automation. I say inexpensive, because you’re not paying for a robot that can do 300 things. You’re building a robot that does one thing very well. And another thing is, it’s self-controlled, meaning you’re running this on your own PLC code. You don’t have to learn a new program to interact and control a new function and automation, as with a off-the-shelf robot.

Griffin Lemaster:

So X-rays and robots. And if you guys haven’t gathered from my enthusiasm, maybe I don’t give it off very well, this job is really cool, right? Working with X-rays and robots. So how do we here at Creative Electron employ these robots? Well, any way we possibly can, any time we can. Why not use them if the potential is out there? So depending on the task, we can utilize any number of robots to fulfill the requirements of the project. Off-the-shelf products or in-house creations, all of these will work.

Griffin Lemaster:

We at Creative Electron, our expertise is in cabinet X-ray inspection. If a customer has a product that they need inspected for process verification, or they need to do a final test quality inspection or even reverse engineering, we can build a system essentially meets the requirements. One of the key drivers that enables us to serve so many customers in the nondestructive field is our flexibility and our ease of which we can deploy new solutions very quick, and robots have been a big part of this. Robotics, automation in general, it’s been a big part of us being able to serve so many clients and automate a inspection process.

Griffin Lemaster:

So inside the X-ray system, we put robots there really because we can’t be there, right? It’s a dangerous environment. Everything in the cabinet is exposed to high levels of radiation from the X-ray, so we need something moved and manipulated within inside of the X-ray system. Robots are ideal for this type of automation. It’s a dangerous environment, we can move at really high speeds, and it’s repetitive motion.

Griffin Lemaster:

So this is a video shows a SCARA type robot inside a customized TruView Prime X-ray system designed for ball inspection. The X-ray source is there on the left, and that big, dark circle is the image [inaudible 00:16:00] for on the right. What the robot does, is it goes and picks up a ball, brings it up, takes an image, rotates it 90 degrees, takes another image, and then basically the computer will analyze those X-ray images, classify the ball, and the robot will take that ball and then drop it off into one of the designated locations for the results of that analysis.

Griffin Lemaster:

Now, the video I’m playing here, you’re watching at about one-third speed, and that’s just so you can actually see everything that’s going on. Through the zoom, I don’t even know with the lag if you can see well, but it’s essentially doing one of those every second, so it’s going very fast. And this customer required that this inspection be done in a circle in a certain cycle time. But even though this movement is, like I said, it’s very moving very fast. It’s small movements. It’s not moving that much. It’s lifting up a couple inches, left or right a couple inches, and that’s it. But with those speeds that it’s working at, it could really still do injury to a person, and so that’s what we have to keep those interlocks, keep it working in a safe condition where humans can’t get in the way of this thing.

Griffin Lemaster:

So this application now shows essentially another version inside an X-ray system. It’s a medical device being X-rayed. It’s a good example of the ugly, right? So it’s a conveyor belt that brings the sample in. Then the device is imaged. The image is analyzed by the software and either rejected or passed based on either operators saying it’s good or bad, or we can actually define these programs to use the software capabilities in-house that we have to write algorithms that will define whether it’s a good or bad part. This one, basically, if it passes, it continues on. If it’s rejected, it gets pushed up the conveyor belt into a locked reject bin further down the line.

Griffin Lemaster:

So as I just mentioned, major part of this automation robotics process is this X-ray analysis that we can perform. So a big part of robotics is obviously feedback, what you get when it gets a signal to tell it to do something else. And so one of our major drivers for feedback is this X-ray and image analysis. In our case, it often comes… Well, after the analysis, you have a classification, right? And so we can do this all in-house using essentially the technology we have to develop AI algorithms. And if you watched the Fireside Chat we had last week, J.J did a great presentation on the AI that we’re developing, and he’s really the expert on that stuff. I’ll just tell what I think it can do, and hopefully it’s right. He’ll correct me if I’m wrong, I’m sure. But essentially every sample that runs through our system, every image we take, builds into this model. And as it more images as you get with more variants, the better performance this model has. It can perform at a higher confidence when it makes those decisions, the more samples it gets and the more displays. And so it’s a cool thing that we’re interacting both these breakthroughs in AI and all these new advances in robotics too really given us a lot of bandwidth and depth of the customers and the markets that we can serve.

Griffin Lemaster:

Here’s an example of a cobot operating outside the system. So this is a simple loading and unloading exercise. This is a rendering, obviously, but as the cobot will be interacting with the automation of the X-ray system, it responds to events happening within the system. So further autonomy would then allow the cobot to make adjustments while placing the sample efficiently. So if it’s getting feedback from the analysis that says it needs to be moved or needs to be twisted or rotated, the robot can actually feedback and make those adjustments on the fly autonomously without people coming in and checking it. Now, this is one of those cobot examples where the operator can just walk up, place the sample tray down, and the robot can load the parts from there. You don’t have to fence it off, or… If this was a robot acting there, you’d essentially have to put a fence or put other forms of automation that would deliver the samples to the robot for placing in the system.

Griffin Lemaster:

Here’s another example, simple example of… This is a video of basically a low-speed application of loading, unloading, and sorting. So after the sample is imaged, it exits one of our custom inline TruView Prime X-ray systems, and the robot will pick it up and sort it based on the analysis. So this is a low speed application, so in that effect, you could use a cobot and replace the robot we have if, for example, you can’t create that fencing, or you can’t prevent human presence from being in that area.

Griffin Lemaster:

And so obviously there’s this common fear that robots will one day take over the world and control all the humans. Well, I can’t predict the future. I really don’t think this is going to be true. From their origins, really robots… Robotic implementation is rarely with the intention to replace human capacity, but instead, really to increase our capabilities, whether that be sorting 1000 boxes in 10 minutes, working in an X-ray chamber, or lifting a 800-degree aluminum casting. But every day essentially, our lives are touched by these new advances in automation, in industrial automation. Science and technology improves. So does this symbiotic relationship between humans and robots.

Griffin Lemaster:

For example, in the medical field, miniaturization has led to crazy breakthroughs in biological, tiny robots. Already microbots are being used to interact with multicellular organisms, and nanobots, or sometimes popularly referred to as nanites, they’re currently being developed to essentially track and trace the movement of substances throughout our body, as far as drug delivery goes and how things work physiologically within our body that we’d never seen before. And even to the fact that… I know it’s not in practice yet, but they’re working on these nanobot cells to replace white blood cells, so for people with immune deficiency issues, they can inject nanites, if you will, and these nanites will be able to actually work on infections and where the human body can’t actually fend for itself. So some really interesting and cool stuff coming out from the medical side.

Griffin Lemaster:

In agriculture, I mean, there’s been robots plowing fields for a while now, autonomous vehicles going around and harvesting crops, but they’re also doing other things. They’re doing sorting of the crops. They’re counting the crops. There’s even weeding robots out there toiling away in the fields. So it’s neat to read about this stuff.

Griffin Lemaster:

In education, I read a very interesting article last week that’s very… It relates to what we’re all going through right now with this pandemic, is that if you have kids, then you’re trying to teach the kids where there’s a lack of a teacher, and it’s really tough. So the funny thing I read in this article about these robots that are actually working on teaching children and teaching them lessons, is that they can repeat the same lesson over and over and over endlessly, and the robot doesn’t get frustrated, and heck, it even encourage the kids. And so they’re pretty neat, what they’re coming up with and what they’re doing there.

Griffin Lemaster:

And so in industry, in industrial automation, where we’re sitting, our sector, we’re always looking for ways to add capabilities. We do that every day, every time we build a new system. So we’re not just adding capacity. I mean, yeah, we want to do things faster and better and do it more, but we won’t all lose our jobs to robots. That’s sort of this myth. I mean, but I can guarantee you, there will always be jobs to fix them. That’s not going away anytime soon.

Griffin Lemaster:

So with that, that’s the brief wrap up of this Fireside Chat, and I’d be happy to answer any questions you guys have.

David Kruidhof:

Awesome. Thanks, Griffin. We just have a couple of minutes. There’s several questions that have come in, but we’ll answer what we can. One of them is, is there any danger to the robots inside of our X-ray machines due to the radiation?

Griffin Lemaster:

So generally not. I mean, the robots we’re using in our systems are for rugged use. They are hardened steel cases. The electronics, most of the electronics that actually control the robots are far outside of the radiation field. You really just have mechanical parts that are moving inside the robot. So no, there’s no danger to the robot, the mechanical parts of the robot. There are very sensitive chips out there and electronics that can be affected by radiation, but if there are some in the controller, the controllers are way outside of the radiation area, so it’s not an issue.

David Kruidhof:

Okay, good. Another one here. Can a cobot pick up items that are in disarray, or do they need to be organized in a bin for it to-

Griffin Lemaster:

Good question. That depends on the capability of your cobot. I mean, there’s several different models out there that can do different things. If you a scatter a bunch of parts on a table, but they’re all the same part, yeah. I mean, obviously there’s not a problem. You can use cameras to find where the parts are and load them. And at the same time, even if they are different parts and you need to pick out the right things, yeah., if you have enough sensors and enough cameras located on the robot, and you have a software that enables it to detect what’s what, or you have some method of feedback to determine where a part is, then yes, it can pick up individual parts from different locations, and it’ll just need a different type of end-of-arm tool to do different tasks, whether it’s a grip or whether it’s suction, so it could use any of those types of methods.

David Kruidhof:

Okay. So it’s really a discussion on a case by case basis.

Griffin Lemaster:

It is. It’s very specific to the task at hand, and there’s hundreds of different end-of-arm tools that they can put on these robots, including in the software that goes into them behind them, that’s a whole ‘nother field. It’s incredible stuff.

David Kruidhof:

Yeah. Of course, the cost factors of the two different options, right? Is it going to cost a lot of time and money to put it in an organized bin, or is it going to cost you a lot of time and money to pick up a disorganized pile, right?

Griffin Lemaster:

Sure. And I mean, obviously as managers we always ask, “Is that going to be more effective if a person did it, or if a human did it? How much do you have to pay them? What’s the risk to them? What’s your liability for having a human doing it versus having a robot do it, and then versus the cost of implementation?” So we all do that cost analysis, and that’s part of our jobs as engineers, to figure out what can be done. If a robot can do it, I mean, yeah, you still have to perform that cost analysis, because yeah, if it’s a very complicated process, it will be very expensive robot and then maybe take a lot of maintenance to keep it running. Most industrial automated robots out there, they’re built rock-solid, right? So if they’re being operated within their limits, they’re going to run for a very long time.

David Kruidhof:

Okay, well, it’s been half an hour, so that’s all the time we’ve got allotted today. Thank you very much, Griffin. Really appreciate this. It’s been really pretty cool to watch and learn more about these robots and cobots.

David Kruidhof:

Well, thank you all for joining us. If you have any more questions, definitely email us at the email here on the screen, info@creativeelectron.com. Look forward to seeing you here next week. We have a good presentation by Dave Phillips on getting the best use out of your legacy X-ray machines, so I hope to see you next week.

Griffin Lemaster:

Thanks everybody.

Speaker 3:

Creative Electron