What's new
What's new

My customer just released a video showing off the robot hand I've been sweating over

implmex

Diamond
Joined
Jun 23, 2002
Location
Vancouver BC Canada
Good morning All:
So I just got an email advising me that the super hush development project I've been busy with has been released for publication and I can show off the video if I want to.

I still cannot release design details of course, and I cannot comment on what I know of it's performance specs.

Even so...it is awfully cool to see...I've linked the video of the uncovered hand.
They have another with a rubber glove on it.
I built virtually all of the mechanical bits for the hands, and made them work, all from un-dimensioned, un-toleranced raw CAD models, so there were a lot of design and build decisions that had to be made from first principles.
There are hundreds of parts, some with tenths tolerances in key places.

I've shown some of the build challenges with individual parts here, over the course of the past year and a half.
Several versions were designed and built, and several more will be designed and built over the next year(s).

I took advantage of every gadget in my arsenal for this project and I had a huge amount of fun doing it.
The company is a great bunch, full of immensely smart young engineers and have been a terrific pleasure to work with.

So here is the link:
https://www.youtube.com/watch?v=FBpAonknMjU

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining
 
Good morning All:
So I just got an email advising me that the super hush development project I've been busy with has been released for publication and I can show off the video if I want to.

I still cannot release design details of course, and I cannot comment on what I know of it's performance specs.

Even so...it is awfully cool to see...I've linked the video of the uncovered hand.
They have another with a rubber glove on it.
I built virtually all of the mechanical bits for the hands, and made them work, all from un-dimensioned, un-toleranced raw CAD models, so there were a lot of design and build decisions that had to be made from first principles.
There are hundreds of parts, some with tenths tolerances in key places.

I've shown some of the build challenges with individual parts here, over the course of the past year and a half.
Several versions were designed and built, and several more will be designed and built over the next year(s).

I took advantage of every gadget in my arsenal for this project and I had a huge amount of fun doing it.
The company is a great bunch, full of immensely smart young engineers and have been a terrific pleasure to work with.

So here is the link:
GPR-1 Hand Control - YouTube

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining

Brilliant, I hope you are proud of yourself.
 
Working yourself right out of a job, let’s build millions & give it enough brain cells to out think a machinist. Maybe we can make it work cheaper than the Chinese ��
 
Hi vonleyser:
Naahhh, it's still a million miles away from even being able to pick its nose with the same dexterity as your 5 year old kid.
Here's another video of it attempting to pick up chess pieces:
GPR-1 Task Planning Demo - Chess - YouTube

Do you feel threatened by it? :D
I certainly don't...not yet anyway.

But it's still an amazing project to have gotten the opportunity to be involved with.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining
 
I built virtually all of the mechanical bits for the hands, and made them work, all from un-dimensioned, un-toleranced raw CAD models, so there were a lot of design and build decisions that had to be made from first principles.
There are hundreds of parts, some with tenths tolerances in key places.
Emphasis added

As an engineer, I cringe. Did they think that barrel was a masseuse table when they bent themselves over it? Assuming they intend to make more in the future.

Great work BTW.
 
I recommend everyone to watch the video fully, as the gloved hand is really "uncanny" and spooky natural. Amazing.

I wonder, if you speak about non-dimension drawing, is the robot hand copy of yours ? :)
 
Working yourself right out of a job, let’s build millions & give it enough brain cells to out think a machinist. Maybe we can make it work cheaper than the Chinese ��

Having worked as a machinist for many years I feel qualified to say that you need to narrow it down a bit. I would say it already has those on the low end of the spectrum beaten.....add a pinch of arrogance and it will call itself a tool maker. Take away some of the social skills it demonstrated and you will have an engineer. :D

All funning aside....very nice work. Well done
 
Hi Labrat:
I actually prefer to build from raw CAD models at these early Alpha stages, mostly because so much of the design is still open at the point where I typically get involved.
There is great freedom in establishing criteria by collectively deciding the end goal, evaluating the subassembly to see what it will take to hit that goal and then try to work to those constraints and see what comes out.
It's very empirical, it's very old fashioned, but it's often so much more effective than endless meetings and arcane arguments.

I have a good sense of what I will need to do, I have a good sense of what I actually CAN do, and I have a good sense of what the implications will be for future production; so I'm usually better positioned to make a useful contribution to that domain than a whole group of clever young engineers who haven't ever really had to build anything before and make it work.

I typically get lots of models that cannot be built as modeled and fit together in CAD with zero clearance, I typically get a quite vague set of performance criteria and I typically get a rough budget within which they wish I can stay.

All is handled transparently so nobody gets an ugly surprise and I take good care to leave the principal design decisions in their hands, focusing only on what I have to do in a purely practical sense.
For example, I know, far better than they ever could, how tight to make a dowel bore so I can still take the assembly apart without beating the life out of it.
What I don't want to decide is how big the dowel needs to be to carry the intended load...they can do that part far better than I can.

An after-action report gives them critical information to decide all the practical details including the tolerances, the datums, the processes, the materials etc etc.
Once out of Alpha, they can pin all those things down...by that time I am substantially out of the picture anyway...my shop is a prototype shop, not a production shop.

Is it a more expensive way forward than a team of young guys, not super familiar with GD&T who know kind of what they want but can't yet put it to print unambiguously?
Actually, no, it's quite a bit LESS expensive, because they can focus their time and resources on the stuff they do well, and I can just build parts to suit from their rough design.

As a practical example...There was way too much for me to build within the desired time frame so we contracted out some of the easiest bits to a well regarded local shop.
I walked the young engineers through all they needed to know, I pored over many of their attempts at unambiguous drawings that actually captured the design intent, and we eventually sent a package out for tender that took far longer to define than to fulfill.

The shop chosen did a good job, but for Alpha prototype parts, it was a very painful exercise: it set us back well over a month and cost more than double my invoice for some comparable other parts I built in under a week just from the raw models, with zero further instructions.
My parts worked at least as well as their parts did.

The customer had to trust me, and to their great credit, they did, so they ended up paying me far less than the project would have cost any other way, and they got it much faster too.
It works well, all are happy, what's not to love!

So for these circumstances, although alarming to an engineer like yourself, it's a viable way forward and it will be superseded very soon by a more conventional approach as more of the details are defined.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining
 
Very nice Marcus. That is impressive dexterity for a robotic hand, a lot better than most I've seen. The articulation looks to be almost an exact match for a human hand. The control of dexterity and agility will come with time. Really cool.
 
Very nice work!

Of course, if the young engineers you work with were really interested in making billions of dollars in pure profit, they'd be inventing body parts other than the hand....
 
Very nice work!

Of course, if the young engineers you work with were really interested in making billions of dollars in pure profit, they'd be inventing body parts other than the hand....

Funny you mention that. I believe Marcus has some experience with that particular market. There were some pictures of a "unique" mold posted maybe a year ago.
 
Hi again All:
My role in this design and build was to be the "hands that make stuff"
I also turned out to be one of the "brains" in the domain I outlined above and it was a great collaboration that is still ongoing as they pick over the weaknesses of the current design and make it better.

I have no idea what the business model is; they are quite legitimately keeping that close to their chests.
There is a good bit of IP being developed here, and I wouldn't be at all surprised to see bits and pieces of it being licensed to various entities working in this space, but I don't know if this is the plan.

There is much much more to this robot than the cool hand construction although I'm very proud to have participated meaningfully in that part of it...there's vision, haptic force feedback, flexible circuit boards and position translation and pressure sensors and all sorts of other goodies that I know nothing about.

Those things may well be state of the art too, and far more important to the project than my bit...I simply don't know.
It's a cutting edge research and development domain, and the company in question has obviously invested heavily, so I presume they have a clear path to reaping benefits from their hard work.
Clearly there is a good bit of investor cash flowing through this effort, and I'm sure it's more than just flash.

So I wish them the best of luck, I hope to continue to participate, and I will keep everyone posted as new things develop that I'm allowed to talk about.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining
 








 
Back
Top