Hi,
I came across this & thought it maybe of some interest to folks on here. Its a research project done by Disney, I am in no way associated just thought you'd like it.
Link to the PDF - Computational Design of Mechanical Characters
Link to Youtube press release - Computational Design of Mechanical Characters Video.
Enjoy
PS can we have something like this please :D
An interesting mechanical character design project
Re: An interesting mechanical character design project
I emailed them about this project around a year ago, and got this response. From what I can tell Disney publishes a lot of papers, but keeps the actual results of the projects in-house. Presumably that's to keep the goodies away from their competitors, wouldn't want Boris and Natisha's spinoff to make a better mechanical Bullwinkle.
Interestingly I didn't close the YouTube window while digging out the email, and it started autoplaying a whole bunch of automaton videos. There's sure an awful lot of them.
OT: I received my Maslowcnc kit last week, and made the trek to the little box lumberyard (if I ever move the distance to the nearest HD is going to be a major factor :) ) after taking out a second mortgage for a couple sheets of ACX; the lumber cost almost as much as the Ridgid router. Wonder if anybody's ever made a 48"/1.2m gear with gearotic, and if the Maslow's accurate enough yet (the firmware is still being tweaked) to make it turn? If Disney'd release that software I have a 30' wide wall to put it on if Mrs. Moose doesn't catch me first.
Kirk
-----
You published a paper, Computational Design of Linkage-Based Characters, on the Disney Research web site in 2014. I'm a member of a hobby gear design forum where we discussed this project in the past, and it was mentioned again today ( gearotic.com/ESW/FavIcons/index.php?topic=1598.0 ). Our membership includes clock and kinetic sculpture makers, and we were hoping that you may have released this software for private, non-commercial, use.
Is this available for hobbyists, and if so how could we obtain a copy?
Thanks!
/s/
-------
Dear Kirk
Thank you for your email.
Unfortunately the paper is not available for public.
Thank you for your understanding and kind regards
Linda Breu
Senior Lab Coordinator
The Walt Disney Company (Switzerland) GmbH
Disney Research Zurich
Stampfenbachstrasse 48
CH ? 8006 Zurich
phone +41 44 632 92 86
fax +41 44 632 15 16
Interestingly I didn't close the YouTube window while digging out the email, and it started autoplaying a whole bunch of automaton videos. There's sure an awful lot of them.
OT: I received my Maslowcnc kit last week, and made the trek to the little box lumberyard (if I ever move the distance to the nearest HD is going to be a major factor :) ) after taking out a second mortgage for a couple sheets of ACX; the lumber cost almost as much as the Ridgid router. Wonder if anybody's ever made a 48"/1.2m gear with gearotic, and if the Maslow's accurate enough yet (the firmware is still being tweaked) to make it turn? If Disney'd release that software I have a 30' wide wall to put it on if Mrs. Moose doesn't catch me first.
Kirk
-----
You published a paper, Computational Design of Linkage-Based Characters, on the Disney Research web site in 2014. I'm a member of a hobby gear design forum where we discussed this project in the past, and it was mentioned again today ( gearotic.com/ESW/FavIcons/index.php?topic=1598.0 ). Our membership includes clock and kinetic sculpture makers, and we were hoping that you may have released this software for private, non-commercial, use.
Is this available for hobbyists, and if so how could we obtain a copy?
Thanks!
/s/
-------
Dear Kirk
Thank you for your email.
Unfortunately the paper is not available for public.
Thank you for your understanding and kind regards
Linda Breu
Senior Lab Coordinator
The Walt Disney Company (Switzerland) GmbH
Disney Research Zurich
Stampfenbachstrasse 48
CH ? 8006 Zurich
phone +41 44 632 92 86
fax +41 44 632 15 16
Last edited by Mooselake on Sun Jul 23, 2017 4:06 am, edited 1 time in total.
Re: An interesting mechanical character design project
I don't remember being able to find the actual paper before (or it could be sometimers...), but it's got a lot of interesting references.
They mention their mechanism library is limited and reduces the number of mechanisms they can build. Maybe Art can work out a trade, Gearotic for this? Start a new project with one of those ArtClones?
Kirk
They mention their mechanism library is limited and reduces the number of mechanisms they can build. Maybe Art can work out a trade, Gearotic for this? Start a new project with one of those ArtClones?
Kirk
Re: An interesting mechanical character design project
Hi:
This paper is a facinating one for sure, I read it occasionally and wonder what I could do
to emulate what it does.. I can picture how to do it. I also see there'd be a lot of work to
do so. At this point Im trying to catch up on promises made, a Cad program for modifying
gears and such, a much better recode of kinematics inside that cad program. Im also doing a
lot of research for future toys and interests, so I don't rule out such a module,
or at least a partial code of something like this at some point, perhaps in the context
of the kinematics re-code.
One of the larger problems with such modules is their 3d nature and interfacing to allow
users to pick the paths in 3d space the end mechanism's have to emulate. The papers and
video's do a great job of showing end results that blow you away, but I know that just like
all the very cool animations on the web from various programs, there is usually a HUGE
amount of work from the user to get to the spot they can hit simulate.
The trick with such things is to make it easy enough that one can quickly get results,
and with 3d representations of the world, that can be hard to do. If its too hard, only a very
few will end up using it, and it fails as a project. Its a tricky balance to achieve, but I do
consider it from time to time as I go..
Art
This paper is a facinating one for sure, I read it occasionally and wonder what I could do
to emulate what it does.. I can picture how to do it. I also see there'd be a lot of work to
do so. At this point Im trying to catch up on promises made, a Cad program for modifying
gears and such, a much better recode of kinematics inside that cad program. Im also doing a
lot of research for future toys and interests, so I don't rule out such a module,
or at least a partial code of something like this at some point, perhaps in the context
of the kinematics re-code.
One of the larger problems with such modules is their 3d nature and interfacing to allow
users to pick the paths in 3d space the end mechanism's have to emulate. The papers and
video's do a great job of showing end results that blow you away, but I know that just like
all the very cool animations on the web from various programs, there is usually a HUGE
amount of work from the user to get to the spot they can hit simulate.
The trick with such things is to make it easy enough that one can quickly get results,
and with 3d representations of the world, that can be hard to do. If its too hard, only a very
few will end up using it, and it fails as a project. Its a tricky balance to achieve, but I do
consider it from time to time as I go..
Art
Who is online
Users browsing this forum: No registered users and 14 guests