Industrial Automation - It Doesn't Have To... Be Manual

Have you been seeing the phrase Robotic Process Automation lately?  Well, you're not alone.  This episode we learn about Robotic Process Automation (RPA) and some of the misconceptions surrounding RPA.

What are the benefits of RPA?

How would one implement RPA?

What are the four reasons to automate, according to Brandon's Brandology?

What does RPA have to do with cobots?

Learn more about Hanwha robots. 

Check out FANUC's robots.

Explore our most recent cobot partner, Elite Robots.

Reach out to us with any questions or future topics!

• LinkedIn:  www.linkedin.com/company/ellitek-inc
• Instagram:  www.instagram.com/ellitek
• Facebook:  www.facebook.com/ellitek
• Twitter:  www.twitter.com/elliTek_Inc/media

If you don't want to click on those links, pick up the phone to call us at (865) 409-1555 ext. 804.

Transcript

Brandon Ellis  0:26  

Hello, everybody, this is Brandon Ellis with elliTek and your host for "Industrial Automation - It Doesn't Have To". And I am here with Beth Elliott, our marketing manager. Hey, Beth.

Beth Elliott  0:36  

Hey, Brandon, we're glad you're back this week. 

Brandon Ellis  0:38  

Yep, I am back. I was not allowed last time; no boys allowed. It was all about the women. 

Beth Elliott  0:43  

That's right. I thank you for Allison for spending time with us last time. Her outlook on being a woman in engineering is refreshing. And her and Julie are just amazing.

Brandon Ellis  0:56  

They are amazing. They work hard for us. I'm tickled that they're on our staff. They do a great job. So yes, thank you, Allison, for sitting in for me, I guess, keeping me from being able to do the thing I enjoy the most. It was very good. And it was a good tribute to last month, which was

Beth Elliott  1:16  

National Women's History Month. 

Brandon Ellis  1:18  

That's right. And so, we wanted to do that as a show of support for that as well. So, thank you, Allison. And certainly, Julie, you weren't on the program, but thank you for your hard work. And Beth, thank you for your hard work here at elliTek. And speaking of which, we have crested 1000 downloads.

Beth Elliott  1:34  

All right. Where's the cheer on that one?

Brandon Ellis  1:38  

Where is it? Let's see. That's close. Nope, that's not it. So, and we're getting close to cresting 1100. 

Beth Elliott  1:47  

That's right. 

Brandon Ellis  1:47  

So, thank you to everyone who has listened, downloaded. I know a lot of you stream and so I don't know that the counter counts that, but we appreciate streaming streamers, and, and all the good comments that we've gotten. So, like continue to share, and subscribe. And we'll keep doing these podcasts. So today, Season Two, Episode Seven.

Beth Elliott  2:09  

Yes. 

Brandon Ellis  2:10  

Go with the title. 

Beth Elliott  2:11  

All right. Today, it's "Industrial Automation - It Doesn't Have To... Be Manual".

Brandon Ellis  2:18  

Manual. 

Beth Elliott  2:19  

Yes. We're going to be talking about Robotic Process Automation or RPA, another acronym.

Brandon Ellis  2:27  

One of the biggest acronym misconceptions in in our engine well in manufacturing in the manufacturing industry, I think because when folks hear robot, and process automation, they think manufacturing floor.

Beth Elliott  2:44  

That's right. That's right. We're going to dispel a couple of those misconceptions today. Yeah.

Brandon Ellis  2:49  

So, let's talk about the definition. And Beth and her wonderful research has gone to

Beth Elliott  2:54  

Wikipedia 

Brandon Ellis  2:55  

Wikipedia. What's the definition according to our friends at Wikipedia?

Beth Elliott  2:59  

Well, according to Wikipedia, RPA is a form of Business Process automation technology, based on metaphorical software robots or bots on or articulated intelligence or I mean, artificial intelligence. Sorry.

Brandon Ellis  3:19  

Articulated, artificial.

Beth Elliott  3:22  

And the artificial intelligence, they also refer to that as digital workers. But 

Brandon Ellis  3:27  

That's AI. 

Beth Elliott  3:27  

Yes, yes, yes. And what the RPA systems, they develop the Action List by watching the user perform a task in the applications graphical user interface, GUI, and then perform the automation by repeating those tasks directly in the GUI. So basically, almost anything that runs on your Windows PC can be accessed by these new RPA products.

Brandon Ellis  3:57  

So too bad for you Mac users out there, according to Wikipedia. It's not gonna work on your Mac. Actually, I see no reason why it couldn't. They're just not developing so much for the Mac. So maybe if you're a Mac developer, this is your time. 

Beth Elliott  4:12  

Hey, so what's Brandon's definition?

Brandon Ellis  4:16  

Brandon's definition of RPA? Well, let me say it this way. It is automation. It is robotic. And it does automate a process. It's a very, let's, how do I say that? It's so many of our users. This is an analogy. Many of our users, I assume, have used a program, a very popular program by Microsoft. It's called Microsoft Excel. And Excel is a spreadsheet software. And so, spreadsheets, meaning you have rows and columns, and you can do a lot of stuff with numbers and text and fields and things of that nature. In Excel, while a lot of people have listened or used it, I'm sure, but a lot of our listeners may not have gone so far as to ever get getting gotten involved with macros or recording macros. And so, a macro is basically where when you record a macro, it watches what you do. And so, with Excel, if you're going to select a section or some data, some numbers, let's say, in a table, and then you might want to sort. Maybe it's one of them is a date field, and you may want to sort by the date. And so, you put your older or newest values first or the older values first something like that. And so you would have a couple of clicks to do that. You have to click and drag to select it. You would then find your, your sort function and tell it to sort by this or sort by that. Maybe you apply a filter, those kinds of things. But your mouse is going to be clicking on menus. Your keyboards going to maybe teach typing stuff in or pressing enter or things of that nature. And the macro recorder in Excel will record every mouse click, every keyboard stroke, and every movement of the mouse as far as selecting places, but it only really comes into play within that spreadsheet, or I'm sorry, yeah, spreadsheet, that spreadsheet within Microsoft Excel.

Beth Elliott  4:36  

okay 

Brandon Ellis  4:38  

I started to go off on database stuff. I've been having meetings this morning with our database guys. And so, I'm back and forth. But anyway, sorry about that. So, spreadsheet is correct term, not data, not data, data table. And so it's really that basic. And so you can play that back. And usually in Excel, you can tie that to a key sequence. So, Ctrl B, or something would trigger this, 

Beth Elliott  6:39  

The sorting action? 

Brandon Ellis  6:40  

The play back 

Beth Elliott  6:41  

Oh, okay, 

Brandon Ellis  6:42  

Of that recording which it's going to do very quickly, mind you. It's going to do exactly what you did every time. So, it is automated. You have to make the key click that, you know, you have to do the somebody, a person has to do the control, whatever. But that's what happens. Well, RPA is essentially the same thing, except it's really controlling your Windows based in this according to Wikipedia, Windows based 

Beth Elliott  7:08  

Applications?

Brandon Ellis  7:09  

All your applications. 

Beth Elliott  7:10  

Okay

Brandon Ellis  7:11  

So going and opening your email client or whatever it is, and opening your email and finding a specific attachment or, or an opening email, and it's got values in there, something like that. And it's got to be repeatable. So, there's one level of challenge there. But if you open up the report, or the email, it's always the same, it looks the same. And the only thing different is maybe the numbers or the names or some kind of text, and you're always selecting in the same place, you know, so many rows from the top or whatever same number of characters, that kind of stuff, I would assume it will mimic those kinds of things. If you copy that, and then you open up your accounting software and, and you hit the same key selections there, and icons and those kinds of things. It will emulate those things for you, in similar fashion. But it's across, not just in Excel, which is macro would be Excel. And I realized that there is even more things you can do with Excel where you can connect to outside things, but it's not really recording mouse movements at that point. But an RPA is really that simplistic. Now, I think that it's a developing industry.

Beth Elliott  8:25  

It is. Yeah from what I've read.

Brandon Ellis  8:26  

Yeah. And it's maybe one day, it's going to have more connectivity, and maybe some of them already do. But right now, as we know, and as Wikipedia knows RPA's. It is a recording type software, which recording your mouse movements and keeping up with what applications you're opening. So that's what I say, that was my very long answer of what RPA is.

Beth Elliott  8:54  

Okay. Okay, there are some misconceptions of RPA. And I think you did, we talked about one of them. 

Brandon Ellis  9:01  

Well, the names are borrowed, I feel like borrowed from the manufacturing floor. So, when you, when you think about robots. Not everybody thinks about the robots you and I may think about because you've seen us install them which is articulated arm robots, SCARA robots, those kind of things, mechanical robots. They probably think about R2-D2 or C-3PO or you know that Will Smith movie where I, Robot or whatever it was called, you know those humanoid style robots and it's called Robotic Process Automation has nothing to do with a humanoid style or a mechanical style robot. Has to do with software-based bots, as you pointed out. That's a big misconception.

Beth Elliott  9:44  

That is. That is. And another one was that it won't completely replace humans. That people will have, I think in my mind, they'll have more meaningful tasks that they can do rather than this repetitive, you know, entering data entry type stuff.

Brandon Ellis  10:00  

Well, you remember in my Brandology, there's four reasons to automate on the manufacturing floor. I think that applies to this as well. If you need quality and consistency, if you want to decrease cycle time or increase production, reduce labor. And by reducing labor, we always said menial task or tasks that are in the manufacturing world if you're assembling or something that could have ergonomic be damaging. So, it's not so much about reducing labor in the manufacturing floor, because you're getting rid of people. It's, it's about not hurting your people. And then there's also the reclassification of people. So, I think that RPA certainly lends to that not so much from an ergonomic Well, maybe I mean, if you count,

Beth Elliott  10:46  

Well, what about carpal tunnel? Yeah,

Brandon Ellis  10:48  

Started from typing at a keyboard. So, I mean, there's that, but mainly, I think it's decreased time. So, most people, most people have plenty to do during the day. In this United States, most of us have not enough hours in the day to complete our task. And that is really the goal is to take the menial task that are always the same. And if you're having to do that, if someone's having to do that, a manager or even an associate, someone in the office, whatever, having to create a report or, or put together some kind of production information or accounting information or whatever, if you're having to do that every day, and it takes 30 minutes to an hour every day. That adds up. 

Beth Elliott  11:34  

It sure does. 

Brandon Ellis  11:35  

And then also, if it's mind numbingly simple, then your mind will numb. And also, most studies, I'm sure have shown that with menial tasks like that, some people hit the nail on the head every single time. But for others of us that aren't very menial task capable because we lose interest; we need newness in what we do, we will begin to become a bit lackluster and pay maybe not pay attention, and then you're gonna start seeing mistakes. Or maybe we do it on purpose and kind of do the Huckleberry Finn and get other people to do. But yeah, for whatever reason, we might not be as consistent. And the quality of our work may not be the best. But primarily, so those two things exactly. quality and consistency. And number two, decrease cycle time just if people have extra hours in the day, they can work on something that's not so menial, and that's just a reallocation of work hours. So that's where I think RPA really, really fits with that. And so, it's not about replacing humans, it's about reducing their workload, or repurposing their workload to something that's that a human has to do. That's not menial. 

Beth Elliott  12:56  

Yeah, yeah. And the other one, I want to circle back to this, the third one after we talk about the stages that people need to take when they implement RPA. Do you want to go over those real quick?

Brandon Ellis  13:08  

The stages? 

Beth Elliott  13:09  

Yeah, cuz I want to circle back to the not limited to enterprise applications.

Brandon Ellis  13:15  

Well, the stages are, basically there's four stages. Planning and that's probably the biggest stage. So you have to decide what you're going to automate. And then what items or objects need to be part of that. So, we talked about the software and all those kinds of things, those are things the email software, the, the accounting software, and that kind of thing. So again, in the example of an email comes in, there's something there, we need to it's an invoice, we need to open the accounting software and do that kind of stuff. And we do the same thing time after time. And so, you've got to identify the objects that are going to be involved with that. So that's not just software's, but what am I going to be dealing with, so an email, and I mentioned just then an attachment. So now, I have to open that attachment. So, what opens the attachment? If it's PDF, maybe it's Adobe, maybe it's something else. So, what things are going to be opening? Because we have to, we don't have to, we just need to make sure that they're available, because it has to open when you click on it.

Beth Elliott  14:17  

Yeah

Brandon Ellis  14:17  

Remember, we're mimicking clicks. 

Beth Elliott  14:19  

That's right. 

Brandon Ellis  14:19  

And then when it opens, it needs to open with some type of consistency, because we're mimicking clicks. So, what we copy and paste has to be kinda in the same place every time. If it opens up with an advertisement or a pop up or something like that, that might, unless that happened when you recorded it, but if it doesn't happen the next time, an RPA software is going to be looking to click on that to close or no thinks. So, if there's got to be some level of consistency, at least as far as I can tell. Not that I'm an RPA Expert. And then finally, you're going to you're going to finalize how you're going to implement all that. So that's all part of planning. Then you have to develop it. Again, if you're using an off the shelf RPA type software's and we aren't we haven't seen one rise to the top yet in my in my opinion. There's a lot out there. There's more coming. It's a developing market, as I said earlier. And then how intelligent is it going to be? What things can it learn? Elon Musk, you talked about artificial intelligence. If you want to really freak you out podcast, not even podcast, just search YouTube or something for some of the interviews with Elon Musk. And he's talking about with machine learning and adaptation with artificial intelligence and how we need to stop, stop, stop, because he paints a picture of something that I only can attribute to, you know, the movie, the Matrix, where the machines take over. And it's kind of crazy. He's a smart guy. He's pretty insightful. But anyway, that's, that's that kind of stuff. And then once you develop it, again, off the shelf software, I don't know, I don't know how that the best way to develop it. Maybe it's off the shelf. Maybe it's actually developing. Developing software is a bit, I think, higher level than RPA is supposed to be. Again, RPA's basic definition is recording what a human would normally do. And writing custom software kind of takes that out for you, at least you hope so. But it also has a lot more time invested. It's a lot larger investment, typically. But still, you're so whether you're recording it, however you going to do that you develop it. And then at that point, it's like, most anything else with software, you have to do a deployment, you have to do testing to make sure that it hits all those little things. So, you know, you go back and figure out a workaround if there is an ad. Or you go ahead and pay the money on the subscription, so it doesn't come back. Something and to get consistency, and then pass that in support and maintenance. So those are the basic planning, development, deployment and then support.

Beth Elliott  16:54  

Okay, yeah. And then on the enterprise side, that the tools that are used to meet the RPA objectives are.

Brandon Ellis  17:02  

Well, enterprise side. So, when we talk about enterprise side, usually we're talking about the IT side from all of our talks about MES systems, Manufacturing Execution Systems. And that's where we our IIoTA comes into play, it connects the OT the Operational Technology, the machines, which are not PC based, most largely to the ERP side or the manufacturing spoke on the ERP, Enterprise Resource Planning, wheel. And so, an ERP system runs the entire organization. That said, we refer to that side ERP, the other accounting, accounting software, shipping, receiving all the PC based software's that are out there in each department are usually on the IT side or the enterprise side. So as far as the needs for that on that side. Again, RPA is more basically recording mouse clicks, I think it's going to have to become more about enterprise and connectivity at some point. Because sometimes your information may not come in an email or something that anybody can open with, you know, PDFs or whatever, it may actually be in a larger database server, database environment, or even cloud-based environment. But even in the cloud-based environment, it's going to be in some type of a data table, or file structure. And so, to be able to go get those files and do stuff with them. An RPA could do that. If the files always in a certain directory, and just like the email. Again, a person it's emulating what a person can do. It's just doing it very quickly and consistently. And so, the trouble is, people are smarter than software. And yeah, anybody. Everybody is smarter than software. Because a human can look at something and say that doesn't look right. Or if, if information is coming in, and it's abbreviated or stuff like that, I mean, goodness gracious.

Beth Elliott  19:01  

That's where is that where the machine learning can come in?

Brandon Ellis  19:04  

Potentially. But again, that's an algorithm, a bit of software that has to learn that. So, it would take a machine a lot longer to learn and recognize that lol means laugh out loud. But for a human, we can sit and think for a minute what does this mean in this context. Because we can put context to it and kind of figure it out. A machine maybe one day, maybe Elon's prediction comes true, and we're all messed up but because machines can figure out what lol means without being told.

Beth Elliott  19:37  

Is that the end of the world?

Brandon Ellis  19:39  

That's it. That's when it all comes to an end.

Beth Elliott  19:43  

With machines

Brandon Ellis  19:46  

When they can use text abbreviations and emojis. So anyway, but the enterprise side is vast.

Beth Elliott  20:00  

It is Yeah. 

Brandon Ellis  20:01  

And so, it, circling back to what you were gonna say it's not limited to enterprise applications. So that's a misconception saying that you have to only use enterprise, to me again IT based solutions. I think we're going toward that. And the truth is so Brandon, how would you do this then manufacturing and OT? And you know, those machines have PLCs, and we don't know how to talk to that, you know, there's no software. We can up with that, you know, the programming software, we're not going to put that in the hands of the accountant or whomever that needs to do this. Well, we have done some stuff, and we can do stuff. It really comes down the organization using the IIoTA and some of the tools that are on our IIoTA. There are other PC based software, SCADA systems and things of that nature that may or may not be able to do this kind of stuff. Usually, we write information directly into a data table. That's really what we're built to do. But we have, we can call on an email server, and we can populate and create an email, and in the body of the email it have information. And we can put that in a very concise and consistent way. We can email that to someone or some RPA bot, and open it, select the information, and go from there. But again, that's one half of the problem, we're sending the information from the plant floor, which is huge for most folks.

Beth Elliott  21:18  

It's the hardest part to do, isn't it? Get the communications in the plant floor? 

Brandon Ellis  21:23  

Well, getting the data from the plant floor in an automatic fashion. RPA, Robotic Process Automation, which is derived from the word automatic, not manual, because Industrial Automation - It Doesn't Have To Be Manual. And so most, a lot, no I don't want to say most, a lot of folks are still on the manufacturing floor, having people use tick sheets and, and write down things and, you know, manually write out or manually enter information that they're seeing reading off of the displays the HMIs, things of that nature of the machines at a certain time, or at an end of shift or that kind of stuff to get that information in. Again, that's just the manufacturing data, we're not going to be in the accounting department with an IIoTA or shipping receiving necessarily, because our IIoTA is built to house native communications drivers to be able to talk to all those PLCs and robots and CNCS and those kinds of things. And so, there's, there's licensing involved with that. And, you know, you're paying an awful lot to not talk to any of the machine. So, it's not really built for that kind of stuff. I'm not saying it couldn't do it, but that's not really where we fit. But if you're looking for manufacturing data to come into and be introduced to an RPA, we could possibly do that through the email. Or again, if they're if the RPA software is higher level, has the higher-level connectivity capability of being able to query data or something like that, then we can push that information into a database for it to query. But again, that goes a little bit higher. That's like RPA 2.0

Beth Elliott  23:02  

There is such a thing.

Brandon Ellis  23:03  

Is there? Maybe they're already there. Can it, can it interpret lol without being told?

Beth Elliott  23:12  

I don't think so

Brandon Ellis  23:15  

There's really an RPA 2.0, right? 

Beth Elliott  23:17  

There is. It's um, I had it on my other piece of paper.

Brandon Ellis  23:22  

RPA 2.0.

Beth Elliott  23:24  

 It's unassisted, RPA or 

Brandon Ellis  23:26  

Where it just happens automatically. 

Beth Elliott  23:28  

Yes.

Brandon Ellis  23:28  

So, we talked about the macros, Okay, that makes sense. We talked about the macros in Excel. And I said, you'd have to have a key sequence. You have to do something a humans got to be there to kind of kick set it off, kick it off. So that's talking about that it would just run automatically. Which scheduled reporting and things of that nature is not far-fetched. So, if there's an RPA software that basically based upon a time, based upon, you know, a calendar, whatever, this month, this if it's end of quarter, that kind of stuff, I can see, see that happening. So, it just needs some means of, of triggering itself. But there again, everything's got to be consistent that it's, that it's reading. So, RPA 3.0. That's where RPAs are able to go out and query databases and pull data back and then incorporate that into your accounting software, you're receiving software or whatever, your manufac even manufacturing, reporting and things of that nature. 

Beth Elliott  24:27  

So, it'd go out automatically and pull it?

Brandon Ellis  24:30  

Yeah. And to be able to have that connectivity. I don't know for sure if they're there yet.

Beth Elliott  24:34  

Okay. Okay. So what tools are available for manufacturers to bring about RPA on their plant floor? Now we're going to talk about robotic process automation. And this is the real robotic process automation.

Brandon Ellis  24:50  

On the plant floor, we talked about the IIoTA as far as being able to pull stuff from manufacturing up and again, that's just one department within the organization. And that the IIoTA is a great option for that, just because it's point and click, there's a very short learning curves and things of that nature, it we tout it at being very simple to use. But if you use robot for what I say it means, and process automation for what I know it for, then suddenly we're not talking about the software we just been sitting here talking about. We're talking about robots.

Beth Elliott  25:22  

Yeah, this is your wheelhouse.

Brandon Ellis  25:24  

Yeah, that's a lot more comfortable for me than then bots. So, what I mean, so think about it, what we're talking about with RPA is, is copying what a human would do mouse clicks, and keyboard punches are whatever. The closest I guess that's what the nature of your question is, what's what would be similar to this on the plant floor on the manufacturing floor that does not have to do with this Robotic Process Automation, RPA that we've just been talking about. But it's similar in what it does. And that would be a collaborative robot. 

Beth Elliott  25:54  

Yeah. 

Brandon Ellis  25:55  

And we are proud to represent plenty of industrial robots. Fanuc is one big one for us. And Fanuc has collaborative options. And they always have for a long time, collaborative robots, but they handle really large stuff. And they can be really, really long handling some really big payloads. And then they've got a new model that is out has been out for about a year but is slowly developing in the United States. And then we are excited to announce that we have picked up an all-collaborative line. They don't do any industrial robots. They've never done that. It's all collaborative. And we're very excited about it. We like the build. I like the fact that it's cast arms and not aluminum tubing. I like the fact that it's got all of the feature sets. I like the interface. So, I'm excited about it, and that's Hanwha robotics. And so, Hanwha has been in the injection molded robot robots for years. And about five or six years ago, they developed a industrial we call industrial, collaborative robot. Now, what's collaborative mean? Collaborative means, what will marketing tell you the collaborative means? Marketing will tell you, you don't have to guard it, you can slide it right in between people and let 'er go. And that's just not true. Everything has to be guarded. Because everything can inflict damage, even other humans. You might not want to slide certain humans in between other two humans, because they may cause damage too. But in this case, in all seriousness, a collaborative robot, we have to look at what it's doing, what the task is, what's the risk of that task, we have to do a risk assessment. It can be stuff like, how fast is it moving? How much range of motion does it have? What's the tooling look like? Is it really large and blunt? Is it heavy? Or is it very light, you know, it can be large and light, or it can be large and heavy. And if it's fairly heavy, and it's moving fairly quickly, then if it hits someone, it could 

Beth Elliott  27:54  

It could hurt.

Brandon Ellis  27:54  

Could hurt them. But if it's light and very pointed and sharp. 

Beth Elliott  28:00  

It could still hurt.

Brandon Ellis  28:01  

It would hurt more. Because even though you don't have the mass there, because mass, you know, force equals mass times acceleration, that's the, that goes back to ole Isaac Newton, is, is one thing. But if you focus that force at a point, so like, if it's an automatic screw driving application, or even a deburr application, things of that nature, repetitive type task, right? That we're trying to get rid of where we want to mimic human motion. That's where collaborators really shine. But safety does have to be, it's more than just it doesn't need guarding. It's got force sensors in there that can pick up on a collision with a person. But again, if I if I collide with you remember, remember your parents teach you don't run with scissors. But they never said anything about running with the Kleenex box. Right? 

Beth Elliott  28:51  

That's right. 

Brandon Ellis  28:51  

So, it's according to what you're carrying, and how fast you're going. Because you could carry scissors as long as you were walking.

Beth Elliott  28:57  

That's right.

Brandon Ellis  28:58  

But you could carry Kleenex box and you could sprint, right? Because it's what you're carrying that comes into play and how fast you're carrying it is really the risk assessment. How much damage can occur, and how quickly can it be arrested. And so those are basic safety guidelines that we use for any robot, any application. But with collaboratives, the marketeers tend to oversell the safetiness of it. And so, I've always said, you don't see a collaborative robot that's not guarded in some way, shape or form. And usually that's by safety light scanners, or maybe light curtains or something like that. But there's some type, it may not be hard guarding, fencing and things of that nature, but light guard, light-based guarding because you want to be able to slow down, arrest, whatever. But that said. Enough about that. Collaborative robots, marketing says you don't have to guard them. You do have to guard them. What Brandon says that that's similar to what I think of when I think of RPA. RPA is mimicking human, human movement. These sensors that are, that are installed in the collaborative robots that will protect humans, they're usually, according to the model, but usually at each joint or each 

Beth Elliott  30:13  

Axis 

Brandon Ellis  30:13  

Axis of motion, and they pick up on force. And so, they can sense when they come up against something that they shouldn't resistance. And so that's for the safety part of it. But also, from a programming standpoint, you can do what's called direct teach, or direct control. So, we can put the robot in a posture, meaning that it's just, it's just floating. And based upon those sensors, if I push it in a direction, the sensors will pick up on that force, and they will give away in that direction. And so, it's, it's almost like a lift assist or counterbalance, except in any direction in 3d space. And so now I can grab the end of the robot and I can move it around. I can move it through a path. I can even say to the operator, take hold of the end of this robot, and perform the task as you would normally perform it. And we can record that motion. There's a little more to programming a robot than just recording a path. This is another thing that the sales and marketing folks come to your site, they show you this thing real quick and say, See, it's done and they sell you this thing. And in a lot for a lot of folks, it's a, it's a, it's an eye opener, because you still have to manage programs within the robot. Now, I really like the interface of the Hanwha for this reason, because it's very intuitive compared to many of the others that I've, all, nearly all of the others that I've come to play with. Fanuc's is pretty good as well. But we've been using the Fanuc or the Hanwha longer, and so we're a bit more familiar with it. But we like a lot of the connectivity that it brings about too. Anyway, I'm getting off base here. Being able to mimic what they do, you can call that from a program and say, do this, run this program, and it will run that pattern in similar fashion. Now, the ability to be able to speed that up a little bit, to smooth it out, those kinds of things, we can get more quality potentially from that. But ultimately, what we're doing is real similar to an RPA. RPA records tasks in a Windows based environment or some type of PC environment. We can record task in 3d environment on the manufacturing floor using cobots. And so that really is, that direct teach is what everyone's after. Because now, if we decide, yeah, that path we taught, needs to be a little different. Most anybody can put it in this mode, and just teach the path again. And then the software is just, the program is just calling whatever path you taught. And so, if you change it, it doesn't care. It's just gonna run that path instead of the path.

Beth Elliott  32:47  

And you don't have to be a programmer to do that, do you?

Brandon Ellis  32:49  

You didn't have to do all the other programs. We can help you with that. But once that's done, now you're empowered with the collaboratives, so that you can do those kinds of things and to reteach those paths.

Beth Elliott  32:59  

Like the RPA software, where any regular person can make it work. It's just like the cobots. Once it's programmed, then the operator can teach it it's path. 

Brandon Ellis  33:12  

Yeah, all the paths and teaching the points and stuff like that becomes very easy to do. I mean, there's a teach pendant involved with any model. But on an industrial robot, the teach pendant is the do all where all. That's all you got. Because you don't have those sensors in the robot. So, if you take an industrial robot, and you put power on it, every joint has a brake, a mechanical brake, just like pressing the brake pedal on your car. So, when you turn power off, those brakes engage to keep it from just falling. Because the motors would just unwind, and it would just fall to the tabletop or the ground or wherever. So, it has to be able to lock itself in position and you cannot move it. I mean, it's got the brakes on. With collaboratives because of these force sensors, it can tell when it's about to fall. And so, it can hold itself in space. It can tell when you're pushing on it and outside forces acting on it. And it'll give way, because it just senses that force and says, Okay, I'm letting go of this position, and it becomes compliant at that point. And so that's because of the feedback or the information that it's getting from these sensors, these force sensors in each joint. And so, any collaborative robot that's out there, has the capability of doing direct teach because they have these sensors in place. Standard industrial robot does not. But the use cases of the two are totally different. Not that you can't mimic what someone does with an industrial robot, but you're going to have to teach that using a pendant, which is almost like a joystick, so to speak, I mean, you're having to drive the robot, and you can't press the buttons in such a way that you can emulate exactly what the humans doing. You're, you're kind of trying to hit points along the way to resemble it, but it's not as smooth as actually watching what the human does and recording it. RPA records, what the human does with the mouse and the keyboard and the applications it comes into contact with. Collaborative Process Automation, I just made that up, CPA.

Beth Elliott  33:49  

Not to be confused with accounting.

Brandon Ellis  35:18  

Collaborative robots are recording those paths. And then of course, you can teach individual points as well. But it can be done in a very easy way by the operator just moving or monitoring what the operator would do if they were normally doing it.

Beth Elliott  35:31  

And then it would monitor based like the vision system?

Brandon Ellis  35:34  

No, you would put it in record, you press record, 

Beth Elliott  35:37  

Oh, wow. Okay, 

Brandon Ellis  35:38  

Just like the RPAs, you're gonna hit record, just like the macro, you hit record and it's like, Okay, I'm watching you. And as soon as you start moving, it starts watching where you go and recording that that path. Now, not all of the platforms out there. Hanwha's not one of them. But not all can, they want, they also record your speed and your accelerations, not just the positions of the path. And the problem with that is, is usually when operators doing this, they're not necessarily doing it at their normal speed. Because they're dragging this robot around. It's not totally weightless, it takes some, you know, you got to move it. And so, to be able to, with the platform, certain platforms, and Hanwha is one of those platforms, part of why I'm so excited, is that we can then take those paths and simply say, go a little faster and increase the speed by a percentage, up to a limit, but we can increase it. And that's one of the things that we have looked for. And we finally found it about, I don't know, a little over a year ago, and have worked with them a lot. And the pandemic slowed everything down. But today, officially, we are the Hanwha distributor for the United States of America.

Beth Elliott  35:50  

Alright. What's the 

Brandon Ellis  36:51  

Yeah, so. Nope. Hanwha robotics. How about that?

Beth Elliott  36:57  

That's better.

Brandon Ellis  36:59  

I need to add the clapping back. We replaced the applause. But anyway, so that's collaborative robots. What do you think?

Beth Elliott  37:09  

I think it's a really nice parallel

Brandon Ellis  37:12  

It is.

Beth Elliott  37:13  

to the RPA software. 

Brandon Ellis  37:14  

Yeah, I think so as well. And so hopefully, it dispels some misconceptions on both sides. On the RPA side, that Robotic Process Automation has anything to do with robotics as far as mechanical robotics or humanoid robotics, and what it is. And how you can possibly if, if you have a situation where you have menial and repeated repeatable task, repetitive tasks that are going on in your workplace right now, Google RPA Robotic Process Automation. And you'll be surprised at what you find. 

Beth Elliott  37:49  

Yeah, yeah. There's a lot of stuff out there.

Brandon Ellis  37:50  

You use one?

Beth Elliott  37:52  

I do. Yeah. The transcription software. 

Brandon Ellis  37:54  

That's right. 

Beth Elliott  37:54  

Yeah. It's, it's interesting.

Brandon Ellis  37:56  

And so, what does it do for you?

Beth Elliott  37:58  

It, I upload the file from our podcasts, and it transcribes it, but I still need to go through it.

Brandon Ellis  38:05  

You have to check it. But what transcribing, but what she's saying is, is it types out, it listens to my East Tennessee accent and hers as well. And it figures out, tries to figure out what words we're saying. 

Beth Elliott  38:19  

Yes. 

Brandon Ellis  38:20  

And it begins to type that out. And so, then she doesn't have to listen and type.

Beth Elliott  38:26  

That's right.

Brandon Ellis  38:28  

More so go back and just check for errors. 

Beth Elliott  38:30  

That's right. Yep. 

Brandon Ellis  38:31  

And so that saves you how much time?

Beth Elliott  38:33  

It saves quite a bit, depending on how Tennessee we're talking. 

Brandon Ellis  38:39  

That's right. And then we talked about AI. And so, an example of AI with that software. Of course, we're going off on another tangent here. But getting off on the first topic. Would be, and I'm doing this by design, because I want to see how your software does. 

Beth Elliott  38:53  

Yes, this will be curious.

Brandon Ellis  38:54  

Is if I say Hey.

Beth Elliott  38:57  

What are you referring to? Hey, Brandon?

Brandon Ellis  39:00  

Yeah. Or, if I say, Hey Beth, how are you? Or, if I say

Beth Elliott  39:03  

There was some hay out in the road.

Brandon Ellis  39:05  

Yeah. Hey, did you see the hay that's out by the road? It's the horse’s hay. But not my hay. Hey. And so, we'll see what that does. And we'll report on that. But that is an example of RPA to a point. I mean, it's very much, you know, specific. It's a software. It's not recording your movements. But it's still an example of something that reduces labor, decreases cycle time, and hopefully gives you some amount of quality and consistency.

Beth Elliott  39:38  

Yes. And it is, it does learn.

Brandon Ellis  39:41  

Oh, it does have AI.

Beth Elliott  39:42  

It does. Yes, yes, it does learn about

Brandon Ellis  39:45  

It's gonna have a. I just gave it a book report. So that's RPA.

Beth Elliott  39:50  

And I think that the same your reasons for automating can apply to both.

Brandon Ellis  39:57  

Yeah, exactly. Again, quality and consistency. Yours might not be as consistent, but it's learning. So, it will become more consistent. And then second of all, decreased cycle time, increased production. It would take you days to type all this out, I think maybe if you type really fast. It would take me weeks, because I type more mistakes than then then that software can possibly create. Reduced labor, it's not that we're knocking Beth out of a job, it's just that Beth has so many other things that she has to do in her daily grind. This is not something she needs to be

Beth Elliott  40:37  

There, there are more meaningful tasks that can be done. 

Brandon Ellis  40:41  

So be able to take, I don't know, one or two days, maybe three days and pull that down to a few hours, is certainly a reduced reduction in the amount of labor that you have to do. Flexibility, quick setup, that's number four, that doesn't really, I think applies much but those first three, I agree with you. So Robotic Process Automation, that's for repetitive, PC based tasks, software-based tasks. We're going to record those kinds of things, and we're going to do what the operator would do, we're going to do it the same way every time. Now we switch gears to Robotic Process Automation.

Beth Elliott  41:17  

Cobotic Process Automation

Brandon Ellis  41:19  

That's on the plant floor. And with collaborative robots, we can teach the same way. We can, we can watch what the operator is doing. And again, this is typically repetitive, maybe ergonomically unsafe tasks, that we can get this, we can get a high-level quality and consistency. It might be able to do it a little faster. It's definitely going to be a reduction in labor. So, they can focus on taking care of other tasks that the robots can't do. But the robot can do this thing over and over. But more importantly, it can save them from carpal tunnel and things of that nature of the repetitive natures of the task. And then with robots, we do have flexibility and quick setup, because we can just change gears, change programs quickly. And so. But yeah, that's a great parallel. And I like the fact that we did, we did this this way.

Beth Elliott  42:09  

Absolutely. Yes. Thank you for your insights.

Brandon Ellis  42:13  

Well, I don't know how insightful I was on the RPA side.

Beth Elliott  42:15  

Oh, well, you're, you know all about the cobots so.

Brandon Ellis  42:18  

With a cobots. I'm totally totally, totally comfortable. RPA is interesting. I really want to see where it goes. I think it is developing. I don't know. I don't know what it means for students go into college now, if developing that type of stuff. I mean, you know, at the end of the day, you're going to be developing probably an application software. It's an application that does all that. But being able to figure out stuff. Being able to learn. I think that's the probably the next step on that. And then this increased enterprise connectivity. And hopefully the IIoTA makes that easier for folks as far as getting their manufacturing stuff. And then collaborative robots, man, they're the craze. Give us a call. We can, we'd love to work with you. And if you need a collaborative robot that's got fantastic capability, that's built well, cast limbs instead of the aluminum hollow tubing, and steel gears, and those kinds of things. But man, just as collaborative and safe as can be with good, good communications capability. And you always hear me talk about communications connectivity, then give us a call here at elliTek.

Beth Elliott  43:29  

865-409-1555.

Brandon Ellis  43:33  

Yeah, that's the phone number. Or you can look us up online www.elliTek.com, elliTek.com. And then we got all the LinkedIn, Twitter's and Facebook's and Instagrams. So pretty much elliTek. So, if you're on LinkedIn, it's elliTek hyphen, I would say dash, dash or hyphen, either one, whichever you, Beth says hyphen, I say dash. elliTek dash Inc, INC. If you're on Twitter, it's elliTek underscore, Inc. And if you're on Facebook, it's just elliTek. And remember elliTek with a k. So, rate, what is it you say all the time?

Beth Elliott  44:11  

Rate n Review

Brandon Ellis  44:14  

Rate and review. And so, we want to have your feedback. We want to hear from you as far as topics that you would like to hear. Information that you would like us to gather and discuss. If you have someone, you'd like to suggest that we consider as a host or to host as a guest on the podcast, let us know that. We have had some folks that have reached out to us and we're talking to them excited about that and getting some more great information out there. So, thank you again, very much to all of our listeners for letting us I'm gonna say eclipse, imminent eclipsing of 1100 downloads, and we want to keep going and keep growing. So, thank you very much for that. Beth.

Beth Elliott  44:57  

Yes, Brandon.

Brandon Ellis  44:58  

I think we've rolled to the end of another episode. 

Beth Elliott  45:00  

All right. Well, thank you for your time, Brandon. Thank you all for listening. We do appreciate you.

Brandon Ellis  45:05  

Alright, have a great day everybody. See you, Beth.

Transcribed by https://otter.ai