Hi guys Philip English this from philipenglish.com. Today, we will have a hands on approach with Inovo Robotics and their robotic arm.
Philip:
So hi guys. My name is Philip English, and I am a robotics enthusiast, reporting on the latest business and application of robotics and automation. My mission is to get you Robot Optimised and to support industry infrastructure and innovation for the next era. Today, we’re here with Henry Wood. So we did an interview with Henry Woods, I think it’s about three or four months ago. And now we’re on site today to actually see the robot live. So we have it behind us, and to just have a look at how this robot works. So I’ll get Henry to introduce himself, can you introduce yourself, Henry, please?
Henry:
Hello again. Hi, my name is Henry Wood and I’m one of the founders of Inovo. We put together Inovo to develop robots that were going to be much more accessible for companies building products in batches. Today Phil’s come down to visit us and I’m going to show him how we program the robot and what kind of things it can do. And really looking forward to giving an overview of the product.
Philip:
Perfect. Thanks for the intro Henry. So the next step guys, is we’re just going to basically have a look at the robot and do a quick light tutorial for you. A bit like an unboxing, but more of just a hands-on Phil. So let’s get into it.
Philip:
We’re going to run through things for you guys and obviously Henry is going to give you an overview and I’m basically just going to hear and ask some questions. And so I hand it to Henry to give us a brief start up of the robot.
Henry:
Sure. Okay, so this is our main product. This is the Modular System. And the really big difference between this robot and other robots is that you can reconfigure it to change the payload and reach. So with most robots, you’ve got to choose a product from the range. They’ve got a five kilo version or a 10 kilo version with 1.3 meters or a 900 mil. And the first thing you’re faced with is this catalog where you’ve got to choose which robot’s the right product for your application. And that’s very difficult for a lot of companies because they’re changing applications all the time. So what we offer is a system where you can change the reach and payload of the robot, depending on what you want to do with it. So what we’ve got here is the arm set up, it’s built out of the three core parts, the wrist, the elbow, and the shoulder.
Henry:
But the link tubes between it are interchangeable to change the reach of the robot. Just as a quick overview, we’ve got a pendant here, which allows simple control of the robot. Say, switching off and on the robot and doing some recoveries, if the robot is left in a funny position. But basically limited to the simple tasks that an operator might do. If you want to program the robot, then you have to plug in a laptop and then you can use all of the advanced features with the comfort of a full QWERTY keyboard and a mouse. Which is a bit easier than the touch screen, in my opinion. So we’ve got a third party gripper fitted here at the moment, this is a Robotiq 2F-85. A really, really nice product. And what it allows is force and speed and aperture control.
Henry:
So you can program how hard it’s gripping something. It could lift up an egg without crushing it, but at the same time, it could lift up a brick or something where you want to apply a lot of force. And it’s really adaptable. So we’ve got some simple controls on the wrist here, including a zero gravity button. If I press that, then the ring goes blue and I can drag the arm around and move it into different positions. And on the interface that you can see behind me, this is giving us a live feed of what the robot’s doing. And it’s showing us the way points we’ve created and the program we’ve built. After that I’ve got direct controls of the gripper, and I can actually add way points to the way guide. So I can build the program really quickly by dragging it to position, recording a way point, drag it to another position, record a way point there.
Philip:
Right. I see, and then when you move the robot to different positions, does it automatically build it in this stack here? Or…
Henry:
Yeah, that’s right. This is basically the workspace. So we’ve got the visualization to see what the robot’s doing and where it’s going. And then this is the program that we’re building. So if we start with a new program and clear that, what you begin with is the start block.
Philip:
Okay.
Henry:
And then we have a toolbox on the left, which is all the different features that’s built into the robot. So I can drag out to the toolbox. I could drag a way point, which is a position the arm moves to. But as you can see, there’s various different bits in there, there’s grip functions. There’s delays, there’s logic, there’s loops, there’s controlling of third party tools. So all these things can basically be pulled out of the toolbox and plugged into a sequence, which is a simple and a logical order.
Henry:
So if I position the robot in a point, maybe put it there. And then I can set from robot. So what that does is that captures the current position of the robot. And I can give it a name, so I could call it 0.8. Or something helpful, this is the door handle. If we’re opening a door, or this is the phone if you’re picking up the phone. And then I can drag it to a new position and I can duplicate this block. And I can set it, give it another name. And again, I can set it from the robot. And it can see in the visualization, here are the two points that I’ve created. And get rid of the grip block in the middle.
Henry:
And we’ve got two points and it will move between those two points. If I play that it will run, moves to point A, then it moves to point B and it stops. If you want to apply more logic, we can use a loop. So we can drag a loop out of the toolbox. Plug that in, and then put those inside it. And now if we play, keeps running around that loop.
Philip:
It’s very nice and easy, quick and simple to use, yeah.
Henry:
Yeah. And the really nice thing about this form of programming is you can’t plug something incompatible into a place it doesn’t fit. So it makes it really intuitive. If it’s clicking together, it’s going to run. If you can’t get the bits to connect, then you’re trying to do something that isn’t possible. So, that you notice it moves in an arc. You could do a linear move, so you just change it so that it basically moves in a perfectly straight line between those points.
Philip:
Oh I see and so you can have a fast bit in the middle, slow bit. [crosstalk 00:00:06:27].
Henry:
Yeah you can control the speed of all of these. If I had a third point on there, if I pause it there. So if I drag another point over here, duplicate that one, call that point C. So now we’ve got a path where it goes there to there to there. And if we run around that loop now, it’s going in a more of a triangular motion. Now that’s still linear. So it’s still doing a perfectly straight line between these points, but it goes to the points and stop. If you’ve got to pick something out of a machine, you want it to do it as smoothly and quickly as possible. So we can take the middle one as having a blend on it. In fact we can put a blend around the whole lot if you want. And it will basically mean that it smoothly moves past those points rather than coming to a stop. So it makes the whole thing run a bit faster. So, if I run that…
Philip:
Instead of having a hard stop at A, B or C, it just…
Henry:
Exactly. So you can see sort of over the B it didn’t stop, it just looped over. And so you can get the whole process running a lot smoother and faster like that.
Philip:
So, when it comes to the end of arm gripper, I saw that there was a Robotiq section in there. So it’s similar to how you program that, you can program that with the robot gripper as well.
Henry:
Yeah, out of the toolbox, we can grab a grip block.
Philip:
And this can be any gripper? [crosstalk 00:07:55].
Henry:
You have to tell it what gripper it is. And what we’re doing, is we’re building libraries for more and more tools. So we’ve been building libraries for Robotiq, OnRobot. We’ve got interfaces that let you drive Schunk. And so, if someone presents us a new gripper we haven’t seen before. The first thing we’ll probably do is hand it over to our engineers and get them to add that library to the driver, so that everyone can use it. Because we want to support as many different tools as possible, really.
Philip:
Perfect, sounds great.
Henry:
So, we’ve also got direct controls on the UI there. So on the far side, on the left there’s a control there where we can basically drag the gripper open and shut. And that lets us control left quite closely, we can set the effort of how gentle it is. So if we grip something like that, it’s gripping it. It’s not gripping it that hard, I can pull it out. If we open it again, and then set the force higher, that’s got it really solid set. So you get really nice, fine control with these, they’re very versatile tools.
Henry:
And one of the things you might notice here is that it’s actually plugging straight into the wrist. It’s not a cable on the outside, and this is one of the limitations with some robots is that you can attach third party tools. But you’ve got to attach them to the control box. One of the reasons we were really keen to have it attached directly to the wrist, is because these joints can just rotate freely in any direction without worrying about a cable getting stretched or snagging. And you could even drive a screw with this and go around 10 times and it will just continually rotate. So there’s a lot of flexibility that comes out of that.
Philip:
So, how does that compare to other sorts, like cobots then? They normally have something you have to plug in and then a wire that comes…
Henry:
They’re starting to offer some support, but it’s quite limited actually. A lot of cobots, you get an interface that plugs into the control box under the table. And then you get a long cable and you’ve got to run five meters of cable around the outside and then you’re cable tying around the joints. And it has just got a lot more chance of actually getting snagged or pulling out, which is a real restriction.
Philip:
Yeah. Sounds very messy if you have that.
Henry:
It can be quite messy, yeah.
Philip:
It’s only be a matter of time before it gets to somewhere it shouldn’t be and then ends up breaking something, so yeah.
Henry:
Absolutely, so you can control the gripper live like that. There’s obviously a block to control the gripper mid sequence. So if we wanted to put a grip in the middle of there and we could say this one’s an open position. And then we could have a short pause and close it again. And then that would basically be running around the loop and it will open at point B and then it will close at the third point after that. So, you can sort of see, you can build up increasingly complex programs in this visual language. Really easily without having to text scripting language or anything.
Philip:
Well, this is it, it’s very, very easy to use, really. Very simple as you just shown me. So yeah, I’m very impressed with that.
Henry:
And then balanced with the complexity, you want to still have fine control. So, if you do need to do something very precise. So if for example, we’ve got to pick this block up. You can can drag it down there and get it roughly in the right position very quickly. But then you might want… I’ll do it this way so the camera can see. Then you might want to get very fine control at the end. So I can stop using the zero gravity mode and I can use the controls on the screen. And then you can do very accurate…
Philip:
Minute.
Henry:
Maybe we can set it to only moves one millimeter at a time. Or you can set it to only move in one axis at a time. So you could move very carefully, even if you’re around delicate things that could be broken.
Philip:
And then if it was something delicate, how do you test? Or is it just a bit of trial and error? [crosstalk 00:11:40]
Henry:
It’s a good question. If you’re figuring out exactly how hard to grip an egg, it could be a bit of trial and error. The point where something will break isn’t something that we can really monitor, but you just gradually increase it. You could start off setting the grip force so that it’s got it. And then you try and pull it out the grip. And if it’s not enough, increase it a bit more. So these motions will let you do Cartesian motions. So down and up in the Z plane, X and Y. And you can set steps so that it will do small movements, or you can do faster movements.
Philip:
No, that’s great. So, we’ve got the controller here, can you just take me through how that works?
Henry:
Sure. Yeah, so this is the pendant. This remains fixed to the robot. So if you’ve got a shop floor environment in a factory, your engineer will probably come down with a laptop and they’ll have their laptop on their knee or the bench while they set up the robot. They will create the program where they can see it with all of these nice features that you get on a full sized screen. So when they leave, they can unplug the laptop and the robot can continue working. This gives a nice, simple interface for the operator to basically set up and run the program. So they don’t build the program on here.
Philip:
So, obviously if you’re in a manufacturing site, you just want to get the robot quickly moving. So you just come to this and uses a very quick, fast interface. And I suppose if you’re doing a different task as well, you can do one task, hit go, move to the next task and hit go. And you haven’t got to go and program the whole robot again.
Henry:
Exactly, the idea of when you come in here, you load program A, and it can do program A for the next few days. And then that finishes and someone’s got another program that the robot’s set up to do. They can just select it from a list here. But what they can’t do, is they can’t edit the program, which is useful because it means that once someone’s built a program here, it can’t be messed with. It’s locked if that’s the only interface you’re using.
Philip:
Right, perfect. So if I… We have a joystick here?
Henry:
That’s right, yeah. The joystick’s quite a nice feature we’ve got there. So, joystick itself would be prone if you just knocked it. So you’ve got these buttons on the side, which are called dead man’s handles. When you squeeze those, it becomes live. So now you’re able to control it. Now you’re in the mode at the moment where you’re basically rotating the wrist around the point. And it’ll only go so far before it reaches a limit. You can jump between other modes. So if we press this button here, then you’re in Cartesian mode now. So, now it’s moving forwards and backwards, left and right. And then doing that still we’ve got up and down. So, that’s basically allowing the robot to travel in this axis.
Philip:
So you can have a full play of different axis’s there.
Henry:
Exactly. And that’s in the base frame. So that means that you’ve got X, Y, and Z like that. So I’ve now put it in the tool frame. So now you’re moving along an axis which is along here. So if you’re trying to insert something, gives you a nice simple way of doing that. That’s quite hard to do with zero gravity because you’re trying to drag it in a perfectly straight line. But in this mode with a joystick, do that. And actually if you move the joystick just a tiny bit, you can move it really quite slowly. So you get a very fine control if you’re trying to do something quite precise.
Philip:
I see, yeah. Because I can imagine if you were trying to do it manually, you could be pulling it in one direction. Not actually know you and not realize you’re pulling to the right, pulling to the left. Whereas here you can do very, very, very precise.
Henry:
Absolutely. So if you’re trying to insert a pin in a hole or something like that, then it gives you the ideal way of doing that. So we’ve got various different ways of doing similar things that suit different applications.
Philip:
Yeah. And that’s a very handy feature to have, well I’d say a brilliant feature to have, really. Especially if this is going to get used in multiple different applications and devices, I can really see that being really helpful.
Henry:
Yeah, we’ve had good feedback on that. And so we’ve got the control to do it with the joystick. There’s an onscreen control and the zero gravity. So typically for almost everything, one of those is the ideal. But there’s no one solution for everything, so it’s useful to be able to jump between them quickly.
Philip:
And then I suppose the next question is, obviously the main thing for this cobot is basically if you’re a small business and you wanted to do multiple tasks. Some of the tasks may be heavy, some of the tasks may be light. So yes, I’d be very keen to interested into how you would make the robot for something heavy and a robot for something light. Can you give us an idea? [crosstalk 00:00:15:59].
Henry:
Sure. Well, this is the super power that this robot offers that others haven’t got. So what we can do is we can change the reach at the moment. This is a medium reach. I’m going to put it upright. Just because it’s a bit easier to handle like that. So I leave it in this position, I look to the screen where I can basically turn off the arms. So, I go, just off. You can hear it click and it’s off. So now what I need to do, is unlock this. No tools, just simple wheel. And then this section comes off.
Philip:
Very nice, very quick and easy.
Henry:
This weighs about eight kilos or something. So it’s not light, but it’s certainly fine for one person to lift and very easy to do on the fly.
Philip:
A one man job, really. So someone [crosstalk 00:16:48] can come along and do it for you, yeah.
Henry:
And the nice thing is you want to move this robot. You can move it in pieces as well. So this is about 10 kilos, about eight kilos, couple of kilos. So all these bits I can move by myself. But if this was all one robot, you’d probably have to get a second person to help with that. So we’ve got a range of different link sections. This is a medium, this is 150 mil. We’ve got a 300 mil or indeed you can put it together with nothing. So if I take this part here, I line up the pins.
Philip:
And this would be the maximum weight that you can get with that extension.
Henry:
Exactly. That’s the trade off. So you’re trading off, reach with pay load. So having put that one on there, turn it back on. See the orange here, blinking indicates that it’s just starting. Takes a few seconds to detect the whole system. And if you look at the user interface here…
Philip:
Oh, there we go, it’s straight in.
Henry:
Popped up, and we can rotate that, so that it’s in the same orientation as the arm.
Philip:
So you’re not there having to download another program to put a different [crosstalk 00:17:49] Yeah, very quick and easy.
Henry:
All we did, was switch off the arm, change the configuration, switch it back on. And we could even run the same program, it’s that flexible. Now I’ll give you a warning if you do that, I won’t do that now. But once the arm’s enabled, I can essentially use it in exactly the same way. It’s updated all of the dynamic systems, the zero gravity and the kinetic. So it’s basically ready to run. We could try and run this program. The limitation is if it’s out of reach. So if you’ve asked it to reach something over here and it’s not long enough anymore. But there you go, so this one’s still valid.
Philip:
It still runs. And then the weight load for this now then?
Henry:
So this is now 10 kilos, so this becomes quite heavy duty in that short reach.
Philip:
Right. Yeah, and then the one bit there…
Henry:
With this, this would be about five or six kilos. And with the longer reach still, then it would still be able to lift three kilos. So you’ve obviously used some of that pay load in the gripper. This is about 800 grams. But at full reach you could still be lifting two and a bit kilos, which is still fairly substantial. And we see a lot of companies where they’re loading machines. They’re putting blank metal [inaudible 00:18:57] in a CNC machine or lifting plastic molded bits up. They really love that longer reach and they often don’t need particularly heavy payloads. So it’s an ideal setup for them.
Philip:
So again, you’ve got three different worlds there, they can use the robot for.
Henry:
Exactly. And actually in the future on a roadmap we’re bringing out an active link as well. So this is a six degree of freedom arm, which is still fairly flexible. Some robots have a seventh degree of freedom and that gives them an extra layer of redundancy. So you can reach inside, something in the elbow can still move. So we’re bringing out on active link section which you can plug in here. And then this is an extra degree of freedom. So you can go from a six DOF to a seventh DOF, which is a really nice feature to be able to have if you need it. But if you don’t have to need it, you don’t need it, then you don’t have to pay for it. Which is our general ethos, you only buy what you need with this robot.
Philip:
So an example of where you would use that seventh degree, can you give an example of that?
Henry:
Yeah, it’s really if the robot’s got to do quite complicated motion. If it’s got to reach inside something, around the back of something, that can be useful. We see seventh DOF being useful in probably a quarter of all the cases. It’s the minority, but it is useful in some spaces. And where we’re finding these longer reaches very useful, is things like inspection as well actually. We’re working on a number of projects, where the robot’s got a camera mounted on the end and it’s reaching around and it’s basically inspecting something from a lot of different angles. Cameras are generally a kilo or less. So it’s an ideal light payload, long reach application.
Philip:
So some of the case studies that you’ve done already, can you talk us through some of those?
Henry:
Sure. Yeah, happy to talk about those. So, we’ve got one of these in a restaurant, a food and preparation space where the robot’s loading raw ingredients into pans and woks. We’ve got another project we’re doing where this is actually a vertical farm and this is loading seedlings into pots. And then they go onto a wrapped conveyor system, where the seedlings go under artificial growing lights. And we’re looking at a couple of projects at the moment where there’s a camera mounted on the arm for all kinds of different applications. In dentistry, in part inspection, in metrology, and in a really wide range of different applications like that.
Henry:
And then the classic applications that people are used to seeing with load robots, like universal robot, where this is an ideal robot for machine tending as well. Because you can set it up for the long reach. It represents very good value if you don’t need those heavy payloads. And it’s also really versatile. If people need to move it to a different application where they might be doing assembly. So, one week it’s reaching inside machine, using the full reach. The next week, got it on a bench and it’s doing some simple assembly task. And then you don’t want that large reach, because it’s just a bulkier system. And you can get better precision and a higher payload in that short reach.
Philip:
Right, so hi guys. So Henry and I have quickly swapped places because this time I’m going to give it a go. Just to show how easy it is to get the robot working. And Henry’s going to give me a hand. So as I understand it, all I need to do, is I hit one button at the top. It turns the LED’s blue and then I’ve got the ability to move the robot handle around. So I’m going to go for these blue points here, this blue block here. Place it down below and then I can just hit this button on the right. And it should give me one position.
Henry:
That’s right, yeah. So it’s added your new way point to the bottom of that list. So if you want to add another one that’s directly above where you are, then it will lift out nice and square. So if you go up again and then… That’s good and then you can just press the button again to add another way point there. Should pop on the bottom of that list there, then you could bring it over this way.
Philip:
So, let’s bring it over here. So we’re going to try and put him on top of this one here, just put it here…
Henry:
Yeah, that’s nice.
Philip:
And then we’ll click it again.
Henry:
One more down there. And then finally you can try and rest it on top of that.
Philip:
Probably about there. I think I got that right, and then one more click.
Henry:
Yeah. So you’ve created now the path for it, and you’ve done all that without having to use the interface. Now, all we do is we’re going to drop a grip position on this. So we can say, so your first point you were starting above it, you’ve moved down over it. This point, you’re closing the gripper and then you’re picking it back up again and you’re leaning over. And if you wanted to drag the 3D, you’d be able to see that that path is drawn at a set point. [crosstalk 00:23:35].
Philip:
Oh I see, so it sits here as well, perfect.
Henry:
If we’re taking a bit more time, what we do is we’d label them each. So we’d call this above, we’d call this over, we’d call this away and call this one place. So that it’s really easy to read.
Philip:
Right, makes sense.
Henry:
But yeah, so now you can go here, you should be able to just press play in the middle of that.
Philip:
Play button here.
Henry:
And it’ll go to the first point. And this is showing you which one is one, so you can see that one’s highlighted at the moment. It’s going towards the next one, sorry. Then the gripper should close, gone up, onto this one. Next one’s going to be the move across.
Philip:
And you can see the way points here as well.
Henry:
Yeah.
Philip:
That’s brilliant. [crosstalk 00:24:12].
Henry:
So now you’ve run it once and it worked. What could we do is we could turn up the speed a bit more and you could gradually tweak it. So if there’s any bits where it’s not perfectly aligned, you can use the on-screen controls to adjust the position and just get it to the point where it’s running really smoothly. And then you can get it to run it faster and optimize the whole process, so that you got it running exactly the way you want.
Philip:
Right, perfect. Yeah, you get that optimization to make sure it’s working perfectly.
Henry:
Exactly.
Philip:
Right, no, that’s great. Yeah, thanks very much for the quick test there. I could see how it’s very easy to move and get going. So thank you for that.
Henry:
It’s a pleasure, it was really nice to have you down here.
Philip:
Yeah, no, that’s fine. Well, thanks for the overview. And as I said, it’s a great product. It’s got many different applications in many different industries, really. So yeah, I’m really looking forward to see how this robot grows within the industry. So thank you Henry.
Henry:
It’s a pleasure, thank you for coming. (silence)
Inovo Robotics Interview – Hands on Inovo Robotics: https://inovorobotics.com/ Philip English: https://philipenglish.com/ Sponsor: Robot Center : http://www.robotcenter.co.uk