Using $300 robot arms and a MacBook to fold a T-shirt(twitter.com)
twitter.com
Using $300 robot arms and a MacBook to fold a T-shirt
https://twitter.com/remicadene/status/1828006677528137977
10 comments
Very cool, many of us here, I'm sure have dreamt of this sort of thing whenever we fold clothes, but you've made it a reality.
Hope to see rapid progress on this!
Hope to see rapid progress on this!
You may benefit from buying a folding board
Yes. Most of the work in folding clothes is in garment positioning. If you can train robot arms to uncrumple a shirt from the wash and position it, you're 90% of the way there.
The folding itself is trivial with a folding board (10% of the work)
https://www.youtube.com/watch?v=AZpkh40AlLE
Even without a folding board, you can do a 2-second pinch
https://www.youtube.com/watch?v=uz6rjbw0ZA0&t=27s
The folding itself is trivial with a folding board (10% of the work)
https://www.youtube.com/watch?v=AZpkh40AlLE
Even without a folding board, you can do a 2-second pinch
https://www.youtube.com/watch?v=uz6rjbw0ZA0&t=27s
I don't know, this looks great for T-shirts but how about for various undergarments, dresses, pants, sweatpants, jeans, etc?
This imitates a human who does a mediocre job of folding shirts. This shirt will end up wrinkled.
I’d be much more interested to see either:
- a robot trained on experienced clothing retail sales associates (can fold a shirt like this in a single movement), or
- a non-human folding technique optimized for the robot arms (having multiple fingers and more arm+hand mobility is part of what makes the human single-move fold possible).
Source: Have worked at Gap, Nordstrom, others…
I’d be much more interested to see either:
- a robot trained on experienced clothing retail sales associates (can fold a shirt like this in a single movement), or
- a non-human folding technique optimized for the robot arms (having multiple fingers and more arm+hand mobility is part of what makes the human single-move fold possible).
Source: Have worked at Gap, Nordstrom, others…
For context, folding clothes is a traditionally difficult problem in robotics you do as a benchmark for how capable your system is. When I worked for a manufacturer less than a decade ago, we were still deciding against doing it at trade shows because the reliability wasn't high enough yet.
The fact that it's hard to get right is the whole point and why the Twitter post is emphasizing that it's being done with very little work on cheap arms.
The fact that it's hard to get right is the whole point and why the Twitter post is emphasizing that it's being done with very little work on cheap arms.
They should first have researched how to quickly fold a t-shirt. It can be done in two seconds with little handling and some practise. I guess even by using only two robot arms. See for example https://youtu.be/uz6rjbw0ZA0
The steps engineers will take to avoid chores! :)
Put it on wheels. Give it a chute for dirty clothes.
I'd happily pay 10x this ($3,000) to never do laundry again if it meant nicely folded clothes would always await me in my closet.
I'd happily pay 10x this ($3,000) to never do laundry again if it meant nicely folded clothes would always await me in my closet.
He has 4 arms there. So $1200 minimum and they look super flakey like they can't dig into a laundry basket.
It's probably easier to build an automated folding table than going through the robotic arm route?
Interesting -- I looked up automated folding tables [1] but even those don't have automatic positioning.
I'm curious: how do they fold with high throughput? I can't imagine shirts are folded manually at industrial scale.
[1] https://www.youtube.com/watch?v=HYVf-1ka36w
I'm curious: how do they fold with high throughput? I can't imagine shirts are folded manually at industrial scale.
[1] https://www.youtube.com/watch?v=HYVf-1ka36w
It's a tech/ research demo
Wouldn't the hard part be grabbing the shirt from the pile and spreading it out flat?
so what limits the speed of the robots? is it computation? or is it mechanics?
motion blur in the cameras?
IIUC, the controller needs to see what pose the arms are in, run inference on that and compute a next step of movement. If there's motion blur, especially with a simple rolling shutter webcam in the macbook, being too fast breaks the vision? That maybe, plus compute I suppose.
IIUC, the controller needs to see what pose the arms are in, run inference on that and compute a next step of movement. If there's motion blur, especially with a simple rolling shutter webcam in the macbook, being too fast breaks the vision? That maybe, plus compute I suppose.
These projects are great!
TIL Dynamixel servos can be had for less than $ALL_YOUR_BUDGET!
[1]: https://github.com/jess-moss/koch-v1-1