Sciencemadness Discussion Board

Robot arm manipulation - for risk aversion with energetic materials

andy1988 - 8-1-2019 at 17:56

Reading these energetic materials threads makes my hair stand on end.

On a project I've been following, sufficiently accurate remote robot arm manipulation looks achievable within an amateur's budget, no programming necessary.

You have two robot arms, one you use any distance away to drive the other arm (the delay in response can be minimized, I forgot their explanation of it), preview picture:

More videos on various models of this arm (as you can see they do a lot of prototyping/research). 3kg payload with a ~1.5m reach.

At present they sell each arm prebuilt for $7400 U.S.D., they are completely open-source, and are much cheaper than other "comparable" commercial robot arms (often 35k+ USD, here is one handling vials). Be skeptical when looking at other expensive imprecise toys marketed for consumers. On thing to be wary of when evaluating robot arms, is the use of stepper motors and limit switches (and a "homing" routine), steps can (and are) lost, especially with resistance and load, leading the arm to be possibly inches away from the intended destination (or maybe even stalled, but thinking it's where it should be)- in the context of stirring a beaker this could mean tipping the beaker over. This arm doesn't suffer from those problems due to the use of the novel optical encoders. But, I suppose the price range may be a bit out of reach if you expect it to be damaged and you don't have a 3D printer to reprint parts, or someone nearby who will do it inexpensively via e.g. library, MakeXYZ, or UPS store. The expensive bits (gears, electronics) are in the base arm, and I imagine in response to feedback they'd be willing to design the option to have the FPGA (~$180) to be some distance away, but perhaps their carbon fiber skin would help shield things sufficiently?

They are working on a cheaper "driving" arm (no motors, gears, etc, just optical encoders). You manipulate the cheaper driver arm and the ~$7400 arm does the same actions.

Other end effectors (grippers) can be adjusted/designed/purchased (if you're into 3D printing), one capable of gripping glass tubes would be all you'd need. As illustrated you can stick the thing in a loop (e.g. stirring a solution with a rod). Maybe mounting the arm on a rail or moving platform if you need that.

They had a cheaper model ~$3-6k USD earlier, but it had a 1kg payload. They've written in the future they expect to bring the price down as they automate their production.

I suppose you'd be especially be interested in building it from scratch if you expect it to get blown to bits. It's maybe ~$1400 USD in materials, but lots of labor/time/glue/expertise, $$ required equipment, and some "supply chain hell" (look through their lengthy assembly manual and webinars).

Unlike other robot arms which seem especially suited to assembly lines, this one seems well suited for impromptu "one-off" tasks by an individual, doesn't require programming, is cheaper, and extensible (completely opensource, others may say they are but in truth are only partially).

All sorts of possibilities with this technology [1][2][3][4], I hope it helps some of you rocketry/mining enthusiasts.

[Edited on 9-1-2019 by andy1988]

DubaiAmateurRocketry - 8-1-2019 at 18:12

Good ideas, something like a robotic arm that is remote controlled for synthesis is something i envision in my future lab :D

DraconicAcid - 8-1-2019 at 18:55

Can I get one that will mark lab reports for me?

DubaiAmateurRocketry - 8-1-2019 at 19:20

But on the other hand, one shouldn't be synthesizing any amount that is enough to blast your fingers off, but it'd still be useful, especially when you do not know the stabilities of newly synthesized EMs

andy1988 - 8-1-2019 at 19:50

Quote: Originally posted by DraconicAcid  
Can I get one that will mark lab reports for me?

Image processing? Would probably require some programming. Might be a bit tricky to handle a stack of papers too, unless you had one of those suction grippers.

This blew my mind, a "universal gripper":

Jamming_Robot_Hand_2.jpg - 51kB

See how it's made!

Oh, more complex tasks could be done with a multi-gripper (2-3 grippers on end, use specific gripper for specific task):

[Edited on 9-1-2019 by andy1988]

Sulaiman - 9-1-2019 at 04:06

I guess the BIG app in chemistry for such an arm would be in a secure location,
operated by remote paying internet users,
but an insignificant app compared to the potential of the porn marketplace.
So as an amteur chemist I'm waiting for cheap robo-porn arms with haptic feedback - that we can modify for chemistry use.