Building a Robotic Fox with 3D Printing, Arduino and Fusion 360
Vložit
- čas přidán 14. 06. 2024
- Here's a new project I've been working on! I've been wanting to make a walking robot ever since my hexapod project so here's a little quadruped robot I built!
My Setup/Recommendations:
🖨️ Big 3D Printer: shareasale.com/r.cfm?b=235472...
🖨️ Affordable 3D Printer: shareasale.com/r.cfm?b=248535...
🔧 Screwdrivers: amzn.to/3WFHXDp
🔌 Power Supply: amzn.to/3UYY6lP
🔥 Soldering Station: amzn.to/3wAYj5u
Support me on Patreon! / nilheimmechatronics
Lethic Z CZcams Channel: / @lethic
Discord Server: / discord
Contact: enquiries@willcogley.com
GitHub: github.com/ikkalebob/NM-bioni...
Videos used/referenced:
- • Watch: MIT’s ‘mini che...
- • BostonDynamics The New...
- • Robot Dog V3 - 3D Prin... - Věda a technologie
Thanks for the mention!
This is the coolest Quadruped Robot design I have seen so far.
Good to see you have kept the body narrow. I see many attempts where the body is wide which makes balance much harder when a foot is off the ground.
Like cats, foxes walk with their four feet almost in line. Maybe the robot's legs can be angled inward to mimic this stance.
Looks nice! I built something similar for my bachelors degree in mechatronics but i like your leg design, our robot had the knee motor mounted inside the thigh, making it quite bulky.
Can recommend MG92b servos, they are slightly larger than the more common mg90 motors but have more torque.
Fantastic!
I'm really pleased that you've taken this project on, and the design so far is really inspiring. I look forward to seeing how you use Unity to establish and experiment with the gait too.
I love this project so much. Such a sleek build and i really like the mechanical setup you went with!
You can use those finger gloves for the feet. Print a "sausage" adapter and slip on of those in place. You can tuck the "glove" into the hole, thus locking it in place. A rubber type off finger glove is excellent for grip.
Optional, you can buy kitchen gloves, cut the fingers off those, and you'll also have the bumps they have on those to grip dishes. Those are even more grippy. I'd recommend using a bent thumb (like you'd press it on a table) as the shape for the feetums. If you want some padding, you can print the "thumb" foot, without the thumbprint area and put some silicone in there, shaped in a negative form, using the glove as a holder (so you don't have to deal with releasing the silicone).
None of the above are my ideas: these are some of the tricks used by the awesome people who make prosthetic limbs for cats in my area. I just print the parts for them as donations from time to time.
Love the way this looks! Good luck with the controls!
Yessss I've always wanted to see proper digitigrade legs on a quad robot!! They always skip that aspect in other projects and I suspect they don't realize just how beneficial the extra joint is! I'm super excited to see where this goes~!
Incredible work, so satisfying to see the design process
WOW!! this is the coolest quadruped design I've seen so far. can't wait to see the next upgraded verion.
Nice, i like the leg mechanism
Every time you upload its inspiring
Woah small world! I was already subbed to Lethic Z because of his insane biped test videos. He had like, 200 subs when I checked last. Cool to see his channel grow - he deserves it!
Neat! I like the clean design. Good luck with the project
he returns
Beautiful design
I love it. Very elegant design!
Very cool very cool
Well done, dude! Look forward to seeing the progress.
Nice work.
Nice work man! Great to see a unique design like this, I'm working on a similar quadruped robot, also good idea placing the motors on the body for a more natural look, think I will try that on my next build.
He's baaack
So happy you’re back dude! You were a big inspiration for me few years back
Lovely, cant wait to see more progress on this
So happy your uploading again
wow i still remember how excited i was for this series back when covid happened. Im so glad its back
The leg structure is very similar to what I wanted to make about 15 years ago. So nice to see that you made my dream come true :)
Will, as you probably already know, there are foam filaments which are used in 3D printed RC planes which you could print the head and tail from to save weight. ColorFabb I believe it is.
LW PLA
Looks like what my dog would do to my pillows. Really nice experiment!
Really impressed I'm looking forward to watching your development process
Ps my timeline is incredibly chaotic too 😅
for the feet - it might serve your purposes to print up a 3-piece mold and cast urethane rubber (look at SmoothOn for good ideas and possible source) slip-on feet/shoes. that would allow ease of replacement for experimenting with the material, and later maintenance.
Great looking design!
that is a neat little unit.
well done man.
reminds me of some Barnaby Dixon animatronics
As far as I remember Dixon doesn't make animatronics... well, one of his creatures has an oled display for face but that doesn't count, he mostly makes hand puppets
This video was my second notification, under Dixons Sinnerman release - small world.
Yeah i was thinking the same thing
Lovehis puppets
tbh, this looks like one of the best base bodys for projects, its jsut lacking coding knoledge,
Good to see you :))
Woooo
Always excited to see one of your videos!!!
Good job, the best inspiration is observe the nature
Really impressive!
This is really cool! I love the simplicity in the legs :) Have you thought about articulating the spine? That would help tremendously with turning.
Well explained video and nice to see some insights and thoughts that you have. I believe that you will make it work of you work hard enough 😉👍. I know what I am talking about because I built and autonomous stair climbing vacuum robot - sometimes it was very complicated but right now it is working 😊
Your finished product looks very nice, but I think that you have ideas what you can to in future. What are this ideas?
i think in these animals, the back arches and plays a vital part in it sprinting. would love to see that being implemented!
Oooh, I want one! There is a product called plastidip, or some such, which is meant to rubberize handles of tools and such. Such a tool might work for rubberizing the feet of a little robot.
A) the design is beautiful and B) at about 7:10 the rapid jitters made me laugh cause it reminded me of a shakey Chihuahua
Very cool and I think you'd probably benefit most from testing different limb section length combinations and of course sticky tabbies for the toes/feet. Perhaps even adding some springs to the "foot" and a possible extra rubbery toes section.
If you get this design working I think it could become very popular as it's very "simple" and clean.
criminally under viewed content! wow! good job man
That's incredible
great mechanical design
Elegant design!
Wow, that's such a cool leg design! I love how compact and slim it looks.
As for controls, I recently played around with Steam Deck as a wireless robotic remote and it was super convenient to use. It has WiFi, Bluetooth and USB for connectivity, huge screen for cameras and/or telemetry, and a set of gamepad controlls. In my project I used a WiFi router with custom firmware as a remote receiver/transmitter, but you can easily use ESP32 instead.
That's a great idea. Especially since it has diverse inputs.
@@AdobadoFantastico For my projects I often use Pi with a screen and xbox 360 controller, and it worked nicely so i decided to build it into a single enclosure. I was about 4 days in when it downed on me that I am just making a crappy Steam Deck.
Certified Pablo the little red fox moment
I didnt expect people start building DIY versions pf this so soon.
Great Job!
Hey, love the videos man!
I asked ChatGPT if tubes of silicone caulking can be used to get a desired grip/traction on parts for a project by thinning it down and bellow is what it spit out:
It's also important to note that thinning silicone caulking may alter its properties, including its adhesive strength and durability. Therefore, thorough testing on representative surfaces is recommended before applying the thinned silicone to the final project. It could also be a cheaper alternative as well in using this method once you've found which combination worked for you?
ChatGPT :
Yes, silicone sold in caulking tubes can be thinned down to create a coating for enhancing the grip factor on parts of a project. However, it's important to note that the thinning process requires careful consideration and the use of appropriate solvents.
Silicone caulking typically comes in a thick paste form, which is not suitable for direct coating applications. To thin down silicone caulking, you would need to add a compatible solvent to reduce its viscosity. The solvent helps break down the silicone and make it more spreadable.
The choice of solvent depends on the type of silicone caulking you have. Common solvents used for thinning silicone include mineral spirits, toluene, xylene, or certain silicone-compatible thinners available on the market. It's crucial to read the manufacturer's guidelines and safety data sheets (SDS) to ensure compatibility and proper handling of the solvents.
Here's a general procedure to thin silicone caulking:
1. Start by selecting a well-ventilated area or use appropriate personal protective equipment (PPE) such as gloves, goggles, and a respirator if necessary.
2. Cut the tip of the caulking tube and squeeze out the desired amount of silicone into a clean container.
3. Gradually add small amounts of the chosen solvent to the silicone while stirring or mixing thoroughly. Add the solvent slowly to control the thinning process and achieve the desired consistency.
4. Continue mixing until the silicone and solvent are well-blended and reach the desired viscosity suitable for coating.
5. Test the thinned silicone on a small area or sample surface to ensure it provides the desired grip and meets your requirements. Adjust the consistency if needed.
Additionally, keep in mind that commercial silicone coatings specifically designed for grip and traction are available on the market. These coatings often have optimized formulations and properties for the intended application, providing reliable results without the need for thinning or customization.
awesome stuff :)
imagine how much these will progress like these ai dogs might be able to provide a happier life for your real dog by being a play mate
For the feet, what would you think about 3D printing a mould, and pouring silicone paws? It would make them grippy, not sticky, and easily cleaned with water.
Looks very good
So cool!
love it!
Wow , that's pretty insane.
To help aid with weight and maneuverability you could solder in thinner wire on the servos
Cada vez que will cogley sube un video, es un dia especial
honestly, rather than printing out and meticulously figuring out a filament "gripper" paw pad, you could just pop down to a hardware store, get some silicone caulking and drip a bit or pinch some onto the bottom and voila, grip achieved. the advantage of this would be absolute rapid prototyping as you can literally just scrape it off and start again, not to mention filament costs and time designing, and my personal opinion-it's just fun to be more physically involved in creation sometimes, add a human element every now and again, ya feel?
This is really cool, makes me want an electronic pet! xD
Super cool! Are you planning to use MPC or RL for the controls?
Wow those legs look so cool
very cool look.
Nice! I'd love to see this project evolve.
Don't know if it's useful but I needed to calibrate the servos on a old project of mine (needed to know max clockwise + anticlockwise position of two SG90 servos) so I used a current shunt resistor to measure when the servos hit stall current. When the servo hits stall current, you know the servo has reached its max (for my project anyway) rotation and you can use those positions within the main function responsible for servo control. Just means that you run a quick calibration at setup and then you're good to go
Thats really smart! Also with a current shunt you could potentially torque limit the motors by measuring and limiting maximum current (by adjusting the voltage), which is something I would like to do. Never thought about using the current sensing as a limit switch though.
How large of a shunt resistor did you use? Measuring the voltage drop might become difficult if the resistance is too small.
@@MangeDT Thanks for the kind words!
The biggest resistors I had (in terms of Wattage) were 1W resistors and my project had two servos, the SG90 draws about 500mA @ 5v (2.5W) but can draw more current if it stalls so I put 4-5 1W resistors in parallel for fear of burning them out (My project tested each servo individually at startup, so I reckoned that accounting for the stall current of one servo, in terms of the wattage of the shunt resistor, was enough). 1 ohm resistance seemed low enough to me and in testing, via the serial plotter in Arduino IDE, you could clearly see the voltage drop (i.e increased current draw) from when I manually stalled the servo, you could figure out the actual values quite easily using Ohms law, but I just used the raw readings of the analogue pin as a relative value without converting it. From there it was just a case of integrating that code into my project and moving from bread board to prototype board.
Hope that's of help to you
Nice!!!
Fantastic design, the movement looks very solid. Do you have any recommendations for learning PCB design?
So good!
One thing I can say is that all the TPU I have printed might have had some flexibility, but none of it was grippy or tacky. I think one would have to cast soft silicone in 3d printed moulds, in order to get true rubbery parts.
+1 to this. I've had good results making grippy feet from simple bathroom silicone sealant. For an extra tip, you can 3d print a very thin mould wall for the silicone attached to the foot, leave a hole to squirt the silicone into, and then cut away the mould wall when the silicone has cured
7:35 ... Cat making biscuits 🤭
the legs on this one are really spectacular. mechanical conduction is a bitch, without pully cables its a really hard thing to work out in your head.
Class!
For the motor control, if you want to use machine learning, I suggest you look into "RMA: Rapid Motor Adaptation for Legged Robots" and "Minimizing Energy Consumption Leads to the Emergence of Gaits in Legged Robots"
Nice!!
Great job Will!
Love wills videos
well project, all my respects. For the moving parts you should specify the legs' actions to being not simultaneously bcs like cats or dogs, all four leg starts moving at different times in one running period. you should add every leg movement a 5ms delay or something else it would act more naturally.
If you decide on using Unity and are using an arduino for control then I would highly recommend the Uduino plug-in for a super easy serial interface.
I love the design, very biological. As such, the forelimbs are stronger to allow for the neck and head you intend to add. 😊 it's working in your favor already.
Ps. I hope you can add your animatronic mouth assembly to it. Hehehe Those teeth will make it look viscous. 😂
you could probably use fusion 360 to make a mold for the toe caps and use silicone to cast a firm but somewhat flexible piece
✔️Fantastic👍✊
Nice, thank you! Idea: internal tail, for "magic balance"?
For the grippy toes, print them in pla and just spray the bottoms with a spray on bed liner, the pla will want to melt a bit due to the dissolved solvents, but if you do very light coats and let it evaporate after like the 3 spray it will stop penetrating the pla. It's fairly grippy.
Another alternative would be plastidip that they use for tool handles, YMMV there I'm not sure how it would react with PLA, with PETG it was fine. I suspect it may melt a lot, but you could mitigate that somehow .... /shrug
it would be a lot smarter to have power distribution on board; umbilical would be much thinner, and you need to do it later for battery power anyway.
Just come across your channel. Great project and a clever leg mechanism!
For the control electronics, have you got a solution already in mind? If not, may I suggest a board I designed called Servo 2040, from Pimoroni. It's a programmable controller that can drive up to 18 servos, read 6 analog sensors, and has built-in voltage and current monitoring.
Super nifty!
So just getting into the CAD side of 3d modeling for 3d printing (lots of Blender time in) , these files are in Fusion 360.
Can you let me know if I understand this right.
I am going to have to download and install Fusion 360 to look at and modify these file?
I can do that, If it has to be that way.
Thanks, This is a cool project and really want to get a look at it, hopefully build it or contribute to it also :)
have you tried using an old gear system from an old pocket watch it is some thing that you shoed do to see if it fixes some of the mechanical problems in your robotic hands it can fix quite a few things that you might be having trouble with that might be able to fix it i have dun a few things with it myself and did the job that i was hoping for so you should try it
It's cool for the simple design. I have an idea of improving the design but not sure it'll work. If you're interested please let me know.
In my experience TPU and ABS have about the same coefficient of friction on metal despite TPU being much softer
Wait, this is the biomimetic hand guy? Kkkkkk. Did that project get concluded? It's been a while since I've seen one of his videos on recommended.
I think I should subscribe
I hope I'm wrong, but those servos look under powered for their purpose. I've build a few hexapods and my earlist one was made with Popsicle sticks and small servos. It could just barely walk whiled tethered. It didn't have a chance of walk once a battery was added.
I'm afraid your quadruped has the same problem with being under powered. I don't think your robot could walk more than a couple minutes with onboard power.
Again, I hope I'm wrong and I look forward to watching your progress.
FWIW, I used a Wii Nunchuck to control one of my latest hexapods. I thought it was fun to have a single handed controller to control the robot.
Good luck and thanks for sharing your latest project with us.
Would love to see a tail that can adapt to stabilise different gaits.
Isn't his just a fancy looking DIY brute-force non-sensing Heath-Kit???
How do you and commercial companies decide whether to use Arduino, microcontroller, raspberry pi, fpga or a combination of those?