[p2p-research] Robot videos and P2P implications (was Re: A thirty year future...)
Paul D. Fernhout
pdfernhout at kurtz-fernhout.com
Thu Nov 12 04:29:47 CET 2009
Michel Bauwens wrote:
> I see a big contradiction between freefall and total robotization, with
> freefall, who's going to invest in total automation?
> so I would add 2 centuries to the robotic prediction, though I'm not at all
> certain that this will occur, I think it's a capitalist fantasy essentially,
> to remove all human contact with making and producing its own livelihood
> (I'm aware of course that leftleaning people have the same vision from
> another angle)
OK, I responded to this once. I'm going to respond again with a longer list
of videos. Most are short (except the Nova one).
"High-Speed Robot Hand Demonstrates Dexterity and Skillful Manipulation"
"Nova: The Great Robot Race"
"DARPA Urban Challenge 2007"
"Home Assistance Robot"
"ASIMO avoids moving obstacles"
"ASIMOs new artificial intelligence. (ASIMO is learning!)"
"IRobot Packbot action!"
"South Korea's Machine Gun Sentry Robot"
"Sentry Robot to Patrol Maine School"
"Merseyside Police helicopter remote control drone"
"The Autonomous Grape-Vine Pruner"
"Robots in warehouse"
"VMS robotic milking"
"Lely Robotic Barn Cleaner"
"Da Vinci Surgical Robot"
"CTC UT-1 ROV Ultra Trencher - Animation"
"Mars Rover Video with Various Music"
"Skoda Giggle: Happy robots"
"a DIY robot mower"
"Make a Reprap Robot Part 1"
There are probably tens of thousands of these videos out there. These are
ones that I see as a representative overview, including robots in all sorts
of areas like home, factory, outdoors, roads, space, ocean, warzone, police,
schools, and hospitals.
More links to sites that aggregate these sorts of things:
This next one, not a video, is a little creepy, but it is an example of how
robots are now doing lots of things you might not think they would:
"Industrial robot hones virtual autopsies""
"The small industrial robot that dominates the room is in many ways much
like any other. A robotic arm smoothly wields grippers and probes - always
accurate and never tired. But rather than working on cars or computers, this
robot is processing human corpses. ... Their virtopsies combine 3D imaging
of a body's surface with a CT scan of its interior anatomy. The result is a
faithful, high-resolution virtual double of the corpse (see diagram). This
double can be used to accurately determine what killed someone. And it's a
more tactful approach: only needle biopsies are used to sample tissues,
leaving a body essentially undamaged."
So, two centuries to the robot revolution? Or yesterday? :-)
But, yes, it may well *never* be "total robotization". But already we are
seeing vast amounts of roboticization. I would think this trend will only
accelerate at companies try to cut costs. You may not see it as much in
older companies where workers resist it or the management does not think
they can get away with it, but you will probably see it in newer companies
-- new car companies, new computer companies, new medical centers, new
factories, new farms, and so on.
One of the driving forces of some of this is also that it costs about
US$750K to keep a US soldier in the field for one year, and sometimes a lot
more for contractors -- so the US military has a big financial incentive to
automate, even ignoring casualties. Even in a "free fall" situation, the
military may just have even more incentive. In the past it was also hard to
get enough "economic volunteer" soldiers. Although, it's true that with the
current US economic disaster, the US military is meeting its recruiting
goals for the first time in many years. Many people are choosing to take a
risk with their health, sanity, and life rather than see their families
suffer with absolute certainty. Still, even when you can get people to sign
up, robots may still be cheaper, overall. Or, like with predator drones, you
may be able to use teleoperated machines, as well as have better supervision
over them from a central location.
So, if you look at some of those videos, you'll see where I am coming from
and what is working right now. None of those videos are conceptual (even the
two animations about real systems, one deep underwater and one on Mars).
They all show working hardware. Granted, some of those are just working in
the lab, and some of those videos are about teleoperated robots. But lab
robots will eventually progress further, and even teleoperation can make a
big difference in our society. Remote telemedicine lets doctors see lots of
patients across the country without anyone traveling; this is especially
important for countries with not very good transportation infrastructure,
like Brazil, where telemedicine instruments, not even robots, but just
cameras and instruments wielded by a local nurse, can make a big difference.
And those applications may be more P2P over time in terms of knowledge
sharing or global service exchange.
So, there really is a spectrum of automation with things going on all across
the spectrum. The things I linked above are for the most part at the far end
of the spectrum. But even things like dishwashers, clothes washing machines,
and construction bulldozers and backhoes are robots of a sort. We just take
them so for granted that we don't notice them anymore.
I think the P2P issues here are in at least a few areas:
* direct P2P issues for teleoperation and shared expertise;
* P2P issues for a commons of information for 3D printing or other robot
* P2P commons development related to open hardware and low level software
related to robotics (perhaps in collaboration with a place like Willow Garage);
* Economic implications of robotics meaning a rethink of our mythological
and physical infrastructure, and how that interacts with P2P;
* Concerns about a robot takeover or other "singularity", either a few
augmented humans launching a takeover assisted by both autonomous and
tele-operated robots (especially military ones), or, alternatively, a robot
takeover or other disaster assisted by "goodlife" (see Saberhagen) human
collaborators, like ones on a "salary"; example: a simultaneous global
takeover of all 3D printers to get them to print deadly assassin robots in
people's homes while people are sleeping;
* P2P aspects of how a global robot AI commons might work, where essentially
all the worlds robots might form one common distributed mind with shared
intelligence (not saying if this is good or bad -- but for example, a
million doctor robots might share medical information on patient outcomes
and all become better and better rapidly, even overnight -- imagine the
Star Trek Voyager Doctor, but ten million of him networked in a P2P way
through a common medical knowledgebase);
* Peer-to-Peer discussions and modeling and tracking of all these trends or
ongoing issues, like mentioned by the Foresight Institute Open Sensor
project, for example.
So, anyway, take a look at some of those videos and let me know if you still
think a robotic revolution of some sort (hopefully a good one for most
people) is still two centuries away. :-) In some ways, I think we have
already had one, it's just hardly anyone has noticed. :-)
More information about the p2presearch