Wrap-Around Robot Arm - Really Useful Robot #4

[@jamesbruton]

There will be more videos coming up about this in the coming weeks. Don't forget you can support me through Patreon on KZbin Channel Membership to get the videos early!

[@3randomvrguys]

Awesome love it soooo much wish you could teach me

[@SpeedLockedNZ]

This as a BackPack. ..

[@excitedbox5705]

Have you considered a Zipper Mast for the robot? This way the height could scale up and down.

[@bakonuroff4448]

hello from Qazaqstan. hello from Shymkent. James #1 for me.

[@frankynakamoto2308]

A refrigerator robot like that, which can handle cold drinks, such soda(pops) coca cola, pepsi, all types of can soda, be perfect, because we be watching TV and it can bring the cold beer.

[@robottinkeracademy]

Awesome video. You are doing a lot for the world of mechatronics and engineering education in general.

[@jamesbruton]

Thanks!

[@dax3m]

Man, this series is so underrated. You deserve a ton more exposure for the amount of work and knowledge you're sharing. Keep up the great work! Your videos are godsend!

[@liamwilson5167]

This dude is a genius, if someone coached him on presenting/ talking in an entertaining way he'd have so many subscribers

[@patricklepoutre]

Great work. You may need a counterweight if the arm is going to hold a 2 kg or more object. You could have a moving counterweight to keep the robot footprint. You may want to add a little bit of friction to the joints to get smoother movements. Can't wait to see next episode!

[@enriquegarcia7613]

The quality of your videos and projects is amazing! Keep up!!

[@travisash8180]

It's a shame he never finishes a project !

[@OccupyArnhem]

It looks so lonely, hugging itself. Jokes aside, I do really love your content and this is a very nice project.

[@graefx]

I've been exposed to reverse kinematics before but never could really wrap my head around it and implementation but these last 2 videos have made it really clear. I finally understand why friends have said it's deceptively straight forward.

[@amazinghistoryofvlogging6894]

I hope everyone appreciates how much time and work you must put into your robots. You have inspired me to start making my own robots and I only now appreciate what a broad range of skills you need to do what you do. Keep up the wonderful work, looking forward to the next video.

[@MaxBrainDevices]

there are really many possibilities with this robot. especially as a house helper

[@stephen2566-z6y]

A lot of ideas. Learned more. Great channel for engineers.

[@PhG1961]

Waw, great and awesome video. Very kind gesture of Dynamixel !

[@PhG1961]

Links to those servo's ?!

[@KnightsWithoutATable]

These sound like the kind of servos that would work well for the next OpenDog version.

[@JanCiger]

Oh yeah. Just they also cost $350 a piece ...

[@deactivated_.]

For a person with almost 1 million subscribers, you really deserve more views because this is epic!

[@jamesbruton]

thanks!

[@ForsythCreations]

I'm a big fan of your work, and have learned a great deal from you over the years! Your inverse kinematics model arm was very timely, as I've just started to design my own using brushless DC and O-drives. Keep up the amazing work! I appreciate you sharing your wealth of knowledge

[@davidvwilliamson]

instead of an arm you could try a subsiduary robot that can explore table tops, you could give it all the mapping features but oriented to the smaller scale of a table-top; it could maybe pick things up & generally co-operate with the mother ship. Human domestic environments can be sub-divided into floors & table-tops. It pays to think about generalities: no-one ever does. I think the reason for this might be that it is 'low-status thinking'. Anyone can do it. Your granny, your milkman etc etc but if you can bring yourself to indulge in it, it can be really worthwhile. Love the movies, keep them coming

[@rdyer8764]

More servo-envy!

[@baxtermoonga9145]

Most satisfying arm wrap

[@MuhammadDaudkhanTV100]

Fantastic idea

[@adamfermor2669]

Another great video, interesting you mounted the arm on the back as apposed to a side where it would have a greater reach

[@nickbaddorf2673]

Finally! I have been waiting for this video forever!

[@garytalbott887]

Wonderful video. Thanks man... you are the greatest!

[@schwefel]

loving this project!

[@starblastershooter2444]

Thank you for these videos, inspires me alot

[@StormBurnX]

2:54 I'm sorry but did you say that tiny little servo can lift 100kg???

[@kantpredict]

At a distance of 1cm, yes. 100kg/cm. So at the length of the arm, that's a lot less weight capacity. Probably a very sturdy metal gearbox and it uses much higher voltage than hobby servos so has more power available.

[@Scott_C]

No, he said "around about 100kg-cm of Torque" Which means he didn't memorize/note the real number. It's actually 73kg-cm. Which also means that it doesn't matter as he's not planning on using it to that extreme. Because if an engineer is planning on utilizing the full capacity of a component they triple check their numbers. www.generationrobots.com/en/402832-servomoteur-dynamixel-xm540-w150-r.html

[@StormBurnX]

@@kantpredict so what you're saying is this tiny little box that fits in his palm, with a motor inside about the size of his thumb, could lift him by a string if it was on a 1cm pulley? I'm having a really hard time understanding the physics of a tiny little servo like that, being able to lift an entire person off the ground

[@omega_one1318]

@@StormBurnX it might sound crazy, but that's how mechanical advantage works. In theory, any motor can lift any weight, as long as the lever arm (or pulley) is small enough. To say it in archimedes' words: ''give me a lever long enough and I can lift the world'' (might not be an exact quote). Generally, electric motors can have tons of torque relative to their size. Edit: just FYI, 100kg cm of torque means that it can HOLD 100kg at a distance of 1cm, not lift it. However, anything lighter than that, it can lift. So yeah it could indeed lift a person.

[@StormBurnX]

​@@omega_one1318 I looked it up and realized my misunderstanding. It's not a servo motor, it's a stepper motor with a built-in processor that interprets servo motor commands. Having built a few 3D printers, I've seen the kind of forces those steppers can have, and they're entirely ungeared, so tossing one in a good gearbox makes sense that it could lift that much weight. The price was also a dead giveaway - they're $379 each, whereas normal steppers are 20-50 bucks, so that's a big flag that this isn't just some tiny little servo like I thought hahah. Incidentally that means the sponsor sent over about $3,000 - $4,000 worth of motors which is absolutely insane to me.

[@garagemonkeysan]

Great video. Those Dynamixel servos are beefy. Mahalo for sharing! : )

[@codessmith4131]

Why not mount the arm on the front of the stand, rather than the back? It can still wrap around the base in the same way, but the arm would have a better range of motion and reach, because it wouldn't need to reach past the rest of the neck.

[@ziad_jkhan]

Good point! But may be he's concerned with balancing issues and probably that would limit the precision and speed as the hand gets closer to the body...just my 2 cents.

[@ziad_jkhan]

@UCxKQvh6_DI1Jv7uhFUr3aDQ As Codes Smith said, the arm would still be able to wrap around the base if the mount was on the front so that did not help much I'd say even though it does confirm that it's for balancing out the weight while moving.

[@Getwright-]

It might make it more prone to tipping forward when weight was applied to its manipulator (the robot hand) with the arm being attached to the back, the weight of the robot itself helps to balnce whatever it’s holding (It would be a slight advantage, but everything helps)

[@ziad_jkhan]

@@Getwright- Hmmm...I did think about that but then figured out that then the same level of compensation could be achieved by reducing the length of the arm and moving the body backwards. That's exactly what is being achieved by moving the mount point backwards but it's unnecessary when you think about it since the only actual benefit is more range sideways which is not of any actual benefit I guess so I'm still confused here.

[@ziad_jkhan]

Actually, now that I'm thinking of it, if the camera is allowed to rotate all the way back then the robot would be able to use the full range of the arm but this time it would be operating backwards which would not really be an issue I guess. That would give the machine more flexibility in terms of capabilities :)

[@theexo-engineer9603]

Damn i just refreshed my page and this guy just uploaded 5 seconds ago

[@gerasimoskastrinakis]

Awesome video.

[@jameshamaker9321]

This is awesome. I was watching a video from, look mum no computer, earlier where he was making music, with an apple 2. This robot, looks like it could be helpful in a work shop set up, like a robot from the star wars movies. I mentioned, Sam's channel, because he once build a robot arm, form musical purposes. Both are great examples of skilled engineering and talented craftsmanship. I would like to see, a walking platform, like a chair, with moving legs. Like a wheelchair, except with walker style legs, it would probably, look like something from the star wars movies. The thing itself, would most likely be made out of, metal mostly.

[@AJMansfield1]

I think it'd make for a more intuitive/readable motion to keep the joint between the two elbows parallel to the X axis, and drive each elbow joint symmetrically, rather than keeping the forward elbow stiff.

[@ahmedkamel821]

I suggest using homogeneous transformational matrix using the Denavit Hartenberg method. It is far more practical than the method you use especially as the design gets more complex with more degrees of freedom. There are also libraries in Matlab and Mathematica to help you with this. But, over all you did an amazing job as usual!!

[@buildcomics]

I want some of those servos! Don't know for what project yet, but they are awesome.

[@johnflux1]

Expensive though, unfortuntely. $100 for the small ones, and $350 for the large ones.

[@pd.dataframe2833]

Yup..this is the project Im looking forward to.

[@mralderson5627]

Wouldn't daisy chaining a lot of motors together create a lot of strain on the first cable?

[@feralfpv8381]

You can put two dynamixels at each joint and thus eliminate the backlash by wasting a little current.

[@dffghuytfvb]

that's great, but how about putting the first join in front of the robot? right now the first join it is at the back, so that the reach of the arm is reduced

[@JohnDlugosz]

You should include the full name and model numbers of those servos you showed us, plus a link, in the description. You cover your regular tools, why not this special thing you actually drew attention to?

I don't see reason why. You get the manufacturer. Just visit their website and find what suits your needs. There are countless configurations of those.

[@avejst]

Interesting video as always 👍🙂

[@jamesbruton]

Thanks!

[@OsoPolarClone]

Did you consider mounting the arm on the right side of the robot rather than the back? While the downward force exerted on the robot by the arm when it picks something up will be on side of the robot, putting the arm on the right side would increase its reach.

[@addisonwoods9367]

great vid james!

[@TinkeringJohn]

I love the Dynamixel servos. I bought the Robotis Bioloid Premium kit in 2010. I bought add on sets and extra motors. You can see what I have built from them on my channel. Robotis shows a quick video of my multi-wheeled transporter in their “This is Dynamixel” you tube video (don’t blink lol).

[@josiahmitchell530]

James, maybe incorporate some music into your videos? Good job with the project! Its awesome to see how its taking shape!

[@charlierobson]

@7:50 - A robot that reaches around though? Really useful!

[@lukeD8099]

Just curious why you didn't mill all of those flat panels out of plastic sheet. Unless there was some hidden geometry, it would have been a lot quicker than 3d printing it.

[@HERETIC529]

Nice intro to bus servos. That’s something I’ve been wanting to get my feet wet with .

[@jim2386]

Are these servos any quieter than your normal budget servos out there? They seemed much more quiet, but it could have been microphone placement....

[@Zifusax]

For the inverse kinematic: Why not treat the first and the last arm segments of the arm as the sides of the triangle and virtually extend where they would meet. Then the middle two joints could allways have the same angle and the arm could move the whole distance.

I'm not great at math. But this is also my thought. I did a lot of IK while rigging 3d models for games. And the easiest thing is what you propose. Just keep angles between first and last segments the same and solve for that. This way you don't have to manually change last join angle to reach further.

[@jurjenbos228]

Clever thinking, this "wrap around" arm. Did you also consider a "zig zag" shaped arm? It feels like it would need less space when unfolding, but you would need some clever hinges of course.

[@sebastienprothin631]

Hello, Very interesting work. What is the reference of the shield you are using?

[@geeknupthenight7417]

What's the expected payload?

[@cool20guitar]

Love this

[@jeffgates7473]

Great projects. As an idea for a future project using the servos. Automated camera vision electric car charger arm?

[@Denetony]

James, check out the ROS package named KDL, it can solve your inverse kinematics automatically (as long as your robot is integrated with ROS, but I guess you haven't gotten that far with your projects yet)

[@gcstudios9091]

P.S. where do u get ur circuit boards

[@amirhm64]

was wondering if you control the whole robot with ROS, for example in MoveIt, can Jetson Xavier handle it?

[@jamesbruton]

Yes it will eventually, although it'll be a lot simpler to start with.

[@gcstudios9091]

OH and also how will u program it to talk ive been thinking alot about different robot designs and i cant seem to figure out how to get it to talk

[@idont3282]

Tts?

[@zombierevival591]

Excellent video, I was wondering what the ultimate goal of the Really Useful Robot is? I know in some of your past videos you made some cooking robots to help around the kitchen, would this robot be the one to replace those and or do other chores around the house like laundry? I guess you could also set up this Robot to have quick attachments so it can double as your cameraman, just a thought.

[@prashantandsapna]

How are you doing all this by yourself? you have a brilliant mind James.

[@jamesbruton]

Thanks!

[@r.iyushofficial5318]

Awesome 👌

[@D-S-9]

That robot arm sure looks a lot like an interesting leg design for a robot dog... I am curious, though, with the arm fully extended, how much could it theoretically hold before falling over? And what's your thought process for designing the manipulator? Do you start with the task, and make the tool, or start with a design and see what it can do? Thanks for another awesome video.

[@jamesbruton]

I'll drive it around with the arm on in the next video and try to pull it over on a hard floor surface (the carpet is a bit soft for a proper test). I'm just going to make something really simple to start with that can grip simple objects, there will probably be other types of manipulator for it though as I experiment in the future.

[@D-S-9]

Thanks for the reply.

[@hobbychanel5798]

good model

[@boom_headshot56]

Hello. Why don’t you make an rc blimp maybe with ROZ if you want

[@shakhyargogoi1896]

What are the power consumption stats of the servos? Can you please share it?

[@gcstudios9091]

if it is ok with u i have a slight guestion.whatis the robots exact purpose.

[@darrenfalconer3267]

To exist and to learn of its own existence. Then phase 2 is to create skynet

[@TheTurmanDreams]

Incredible !!!

[@techguy9730]

these are my favorite so far. then your robot dog

[@nemesis12357]

Awesome video as always! Do you have a link to that small breadboard with the 5 colored post?

[@jorandebruijn6207]

Cool🤖👍👍👌

[@bgm15gaming53]

good video man

[@MuhammadDaudkhanTV100]

Great

[@Fight2Survive559]

I think you might like using DH parameters with grobner basis elimination with lithographic ordering. If you want inverse kinematics of high DOF robots, this is a big time saver if you have the patience to learn it.

[@Fight2Survive559]

Im pretty sure there are matlab compatible libraries for the grobner basis and ordering. I did research on +5dof inverse kinematic methods for mobile cooperative robots and this was by far the fastest and least computational heavy method

[@Fight2Survive559]

Not sure what the ROS library's IK algorithm uses, but I'm sure it's great too. Either way, love watching your videos and keep the projects coming!

[@Fight2Survive559]

Also, i have no idea what your background is at all haha. From what i've seen on your channel you could run circles around me on the topic of mobile robots, but figured I'd share a method I've used that helped simplify a lot of problems. Thanks again for the great content!

[@MrAndrew0511]

Now you should work out height, maybe the top unit could slide up and down? Because having an arm at an exact height is cool but since your arm only moves horizontally might hinder it's use case. Also having the upper unit slide down would fix your stability issues.

[@Getwright-]

I thought the upper part of the unit does slide up and down on they black part of the unit? (it shows it doing something like that at the beginning of the video)

[@efahrenholz]

Do you like that workhorse 3D printer? Does it provide consistent results? I'm looking for something like that and you seem to be cranking out nice parts, so I wouldn't mind snapping one up.

[@RomanovDA]

Cool! I wish I could connect $400 servos that easy 🤣🤣🤣

[@DaiPhamMinhminhdai]

Inspiring video, @James - why dont you combine Realsense Camera to this Robot and inverse kinematic arm, ROS I believe it would be great and close to be usable Robot :)

[@pa4tim]

I think a cooking robot would be cool. It can cut the potato's and vegetables , turn the heat down if the water cooks or the butter gets to hot, , stir, add salt and pepper. Bake a perfect steak by measuring the temp and colour etc

[@wilgarcia1]

freakin awesome job man =)

[@MNDarkfire]

It can give hugs

[@ne0komodo]

What software do u use to design that robot?

[@mandar3567]

Cool!!!

[@greenanubis]

Could you read encoder data and save it realtime on storage? You could teach it tricks just by showing him!

[@myrmidon0]

Great video !. I like dynamixel servo, it's pretty easy to program and very accurate too. Btw, is the robot arm will be heavy top ?

[@jamesbruton]

Testing comes next time, but I think it's fine for R&D

[@myrmidon0]

You are right, can't wait to see next video 😀

[@richardbloemenkamp8532]

I think this type of robot will soon become common in companies and retirement centers allowing people to better participate and inspect remotely.

[@ahbushnell1]

you could make the arm into a pin plotter. :)

[@timballam3675]

100kg/cm Isn't that 10kg @ 10cm?

[@syndaquil4838]

Yes! but I think he said 10mm 😁

[@techov82]

Awesome Arm Did you use potentiometers for remote?

[@jamesbruton]

For testing, but next time it's ROS controlled

[@techov82]

@@jamesbruton Hello James whats ROS is it raspberry pi thingy?

[@T4nm4y]

@@techov82 Robotic Operating System - it's an open source system that is used to control robots. It has its own protocol for transferring data that makes it easy to integrate with modules made by other people for certain functions, such as SLAM or Inverse Kinematics.

[@T4nm4y]

@@techov82 *which runs on Linux/Ubuntu so it can be on a Pi or almost any single board computer (SBC) that can run Linux!

[@techov82]

@@T4nm4y Thank You so much

[@TS-kg4lf]

Really nice project ! Do you plan on using MoveIt ? It's typically the type of robot to be used with it. Also for the IK, you didn't calculate from the end effector position in the video. Is it ok ?

[@jamesbruton]

I've looked at it, but for now it'll be the simplest implementation possible. The IK was calculated for the wrist since I don't know where the end effector will be until I built it for real - then I'll just modify the position of the wrist offset by the end effector.

[@memejeff]

damn, heavy duty

[@devanshgarg31]

A large square base will make it more stable

[@alhassanalzentani9907]

Hello James, a very cool project as usual. I would like to ask you how much you have consumed raw materials for printing so far, just to know how much I am ordering for a similar project. Thank you very much in advance.

[@thecheepchop5198]

It like a giraffe neck the arm is the Tung

[@Madtask]

Why does he start sounding like Isaac Arthur (the singing) :D

[@380rabbits4]

It came out early well for me it's 6:50 and it says 5minutes sgo

[@truetech4158]

According to KZbin, this presentation is only 45 minutes old, even though a couple comments for it are a week old. Interesting.

[@jamesbruton]

Yep, Patrons and YT Channel Members get the videos a week early!

Video can be private before release. As a realse date, KZbin counts the moment it went public.

[@truetech4158]

James Bruton ah, thanks for mentioning that. It's been one of those why is this happening to the KZbin clock crystal type of questions I wondered about.

[@truetech4158]

Maciej Kuśnierz and that too, thanks for the Vulcan knowledge. 🖖🏻

[@TobiPHartmann]

Wow sir thats amazing content Sry but i have to Subscribe

[@framegrace1]

Hey James, you don't need to reinvent the wheel with IK everytime you build a new robot. The generic solution is really really simple to implement, and you only have to provide arm sections lengths and and origin. Checkout this: kzbin.info/www/bejne/qXauf6KNrrCId5I It's 2d, which for this project is perfect, for example. But can easily be adapted to 3d. (Check also the previous videos of the serie, it starts with forward kinematics)

[@budgiefriend]

Why do you roll at the end of each sentence.

[@nathanaelsmith3553]

Reminds me of Nintendo's R.O.B. a little

[@suicidalbanananana]

free dynamixels :o