According to a CNEWS
story on the recent International Symposium on Mine Mechanization
and Automation we are decades away from fully autonomous mining robots.
Marvin Minsky, who
addressed the conference, said that while it was possible such machines
could be built, his forecast was gloomy: "There isn't anything in the
laboratory right now that is very promising". He notes that one of
the problems is that companies today are financially incapable of doing
long term research and development because it doesn't turn a fast profit
for the stockholders. On the upside, tele-operated mining equipment is
feasible and may be a good intermediate step toward full autonomy.
It probably hasn't happened yet because people in the mining industry
aren't very tech savvy, and there may also be trade unions or industry
regulations placing restrictions upon exactly how things can be done.
As far as mining goes I doubt that commonsense reasoning has much to do
with the problem. For a tele-operated system one of the main problems
will be getting the signal from a person sat at a desk to a robot deep
underground. In underground situations there may be plenty of
obstructions blocking radio communications, requiring some kind of
umbilical cable to be trailed around.
From working in robotic EOD circle for a while I would say that
teleoperated robots are feasible for mining. My old next door
neighbour actually did some work on teleoperated mining robots in the
1960's and 1970's he told me about it. I would say though that as with
all things its probably a question of economics viability more than
anything. Underground mining is at least as dangerous as humanitarian
explosive demining, but I guess the cost of a human underground miner
is less than that of a humanitarian deminer to a company. Its horrible
to think like that but Im sure thats whats holding back development
I think the only place robot miners will go is to the moon or asteroid
or somewhere out there where humans can't stay. As dangerous a place
as a mine shaft is, a robot would have to be smarter than a human for
humans to trust it not to cause a cave-in (or drive a car, but I
digress). I don't see that happening anytime soon. So dumb robots
would have to be teleoperated and that doesn't free up or replace a
miner. Robot teleoperated miners on earth are certainly technically
possible, but why would a company spend millions developing or buying
a robot that doesn't replace a miner or save them any money? The
robot still has to be teleoperated by someone. Mining companies
already have mining equipment which would be cheaper to use than
letting the miners have the luxury to telecommute. So I guess you
could say It's not really the robotic hardware technology that's
holding mining robots back (teleoperating), although, if software
brain technology was there, there'd be a lot of robots around not only
in mining but everywhere else. AI4U can't program & prove his
imagined ideas, therefore so far no one has a proven robotic brain
that thinks for itself. And, therefore all we do is piddle with
robots that bump into walls or hard-codedly (aka dumbly) pick up cans
at best. ...still waiting for an AI breakthrough.
There is a lot of point in teleoperated mining machinery. It means that
you can optimise the workplace and the coalface separately. A human
being needs oxygen and neat stuff like that. The robot miner doesn't.
Also, what does it matter if an autonomous robot does cause a cave-in?
You can dig them out.
The real answer as to why we won't get autonomous mining robots on Earth
for some time is that children are cheaper and oh so much more
expendable. From the big bosses point of view why replace cheap
labour with an expensive machine and a highly trained technician? To
save lives? Ha! </marxist> ;)