Top
Best
New

Posted by nharada 5 hours ago

Stringman, an $800 open-source room scale gripper robot(neufangled.com)
188 points | 90 commentspage 2
BHSPitMonkey 4 hours ago|
I love the ingenuity, though it's all fun and games until someone or something manages to yank down on one of the cables...
HeyLaughingBoy 3 hours ago|
Probably sell more of them as expensive cat toys.
passive 2 hours ago||
Very cool, might have to try to rig something like this up in the workshop side of our basement, which is almost always a mess.

$800 is already super cheap, and the openness makes it likely you could implement even cheaper options yourself.

HeyLaughingBoy 3 hours ago||
I was actually just thinking about something like this this morning. Mainly because I was idly wondering if the kinematics would work with just 3 supporting cables. Intuitively, it seems like it should, but I haven't bothered to research it yet.

Anyone?

wave_function 3 hours ago||
3 cables should work. Each cable length is a single parameter that defines a sphere around the mount, and the intersection of three spheres defines a point.

I suspect the reason they chose 4 is that 3 cables would limit you to a triangular region, so you’d only be able to access half the room.

swid 2 hours ago|||
You can join two of the cables together that come from each corner unit into one. each unit - still with an eye hole in another corner - moves the cable left or right. With the opposite corner unit, you can pick X,Y on the ceiling. I'd put the third cable in the grabber, which can be used to lower it to the floor.

Probably if you went this way, you'd want to measure each ceiling cable so it is about X+Y length, or otherwise the slack needs to be held somewhere.

This is one less actuator than their system if I understand them, and they are not counting the gripper itself.

komodo99 2 hours ago|||
Me, i'm wondering if the coreXY kinematics could be adapted to make do with only two actuators on four anchors.
jameshart 3 hours ago||
Range of motion would be confined to the space beneath the triangle bounded by the three anchor points. Most spaces are rectangular, so bounding it to four corners would seem likely to give a more useful range of motion.
stickfigure 3 hours ago||
Perfect for your triangular rooms!
koolba 2 hours ago||
This is really cool!

Would more taut crossed wires and a descending line for the grabber (like the one at an arcade) solve the humans crossing wire path problem? Then it becomes separate x/y and z.

demritocracy 1 hour ago|
A lot of rooms have fans and light fixtures too, another constraint to account for. I picture the wires taut and as close to the ceiling as possible
delichon 2 hours ago||
They compare it to a robotic arm, but they are a natural combination. Two arms, even, floating above it all, could scratch the dog, make breakfast, and maybe someday, fold a fitted sheet.
stronglikedan 2 hours ago|
Even if all it can do is scratch my back, I'm in!
jayd16 3 hours ago||
Can it make a bed? Actually, can it lift a comforter? Can it fold towels?

I think the form factor is actually fine but the specs probably need a boost.

monkmartinez 3 hours ago|
These things are scalable... but I am not sure the PAF "partner acceptance factor" would pass depending on what you wanted to accomplish.
stickfigure 3 hours ago||
Realistically, this has one application: The kids' room. PAF might be higher than you think.
swid 2 hours ago||
They designed the cables to go down into the room, instead of from the ceiling - seems like that is a deal breaker for most environments actually, because you need the whole room to have open space just to drop it to the floor all over.

They say they use 4 actuators, but if they only used 3, this problem would be solved.

Instead of independently controlling all four lines, you can join the lines leaving each corner unit and then just move the line left or right instead of reeling two different lines in and out. You still use the corner eye loops, and you use two cables that are X+Y in length. This allows you to pick a point on the ceiling. Then a third actuator and line in the gripper itself could lower itself to the floor.

cloudfudge 2 hours ago||
If I understand what you're suggesting, wouldn't this result in a lot more swinging since the gripper would be suspended from a single point directly above it?
swid 2 hours ago|||
You seem to. That is a good point I hadn't considered - the tension from the sides holds it in place. Maybe we need to use a telescoping arm (which doesn't seem practical, I'm mostly joking). It might still work, but I imagine if you don't grab it on the first try, you may have to go back to the ceiling. And hope there are no strong drafts in the room.

EDIT: maybe lower it on three lines instead of one?

CyLith 2 hours ago|||
That too is a solvable problem if you have the sensors for it. You can critically damp or choose a motion profile to have essentially no swing.
carb 2 hours ago||
How do you prevent having to grab things perfectly in their center of mass? (Because that requirement seems limiting)

Rigid objects would result in the gripper twisting, but the current design could offset some of that by pulling harder in one direction?

swid 2 hours ago||
With the video, it is casually dropping things like clothes and toys into a bucket, so I'm not sure it would be a big deal. But you could also lower it via three lines instead of one, so there is something to counteract that.
scirob 2 hours ago||
This guy does well on Instagram. Good hustle hope it works in irk reviews. Hope it survives to get to v3
MisterTea 3 hours ago||
I've been wanting to build a cable robot for Halloween which flys a ghost around the front yard with a vision system that detects people passing by and the ghost floats towards them and follows them.
darkmighty 3 hours ago|
I didn't understand if exactly this feature is available, but I think it would be a good idea to let the strings actually come off each motor completely when yanked hard. Interesting idea :)
More comments...