Originally posted by 29083010384959
That's your opinion the progress isn't that big, I happen to think in just the last 15 years AI has grown in leaps and bound, take the
DARPA grand challange, having a rover that could be given autonomous commands to carry
out by a mother satalite that were pre-programmed by our engineers would reduce having to control and look after more mundane aspects of the mission,
such as software the could automatically get the rover from point A-B without having a human behind a computer navagating around every rock, day in
and out untill they reach the location.
Yes, that is my opinion, the opinion of someone that has been following (although not as close as I
wanted) AI for the last 10 or 12 years.
If you look at the end of that page about the DARPA grand challenge you will see that this was not a great leap from the 1994
VaMP, and in simulators that type of challenge has been carried out since the start of AI.
What has really changed is that the hardware is much faster and smaller, the sensors are much better and they used GPS to navigate on the DARPA grand
challenge instead of using just radar and imaging as the VaMP.
There are also cases of rovers doing much better than we expected such as the Spirit and Oppertunity rovers which were designed to last for 90
sols and ended up lasting 15x longer than that.
One of the reasons for those rover's success is the fact that they had their software updated
several times.
Yes, but we would be able to share what we see with our crew mates,
That's not good enough, we need the data to study things, not the
opinion of someone about what he/she saw, and that would be what we would get in a situation like that, it's very hard to be completely neutral while
describing something we have never seen, and what we may consider more interesting in some "scene" could be considered unimportant by other people
that would find more interesting something to which we did not even looked a second time.
and use dedicated cutting edge camera technology to capture the images and send back to ground control instead of being stuck with lower
resolution cameras that probes are usually equipped with due to design restrictions.
That is the problem, design restrictions.
If you compare the restrictions between a robotic mission and a manned mission you can see that if we use just half of the restriction of a manned
mission on a robotic mission we would be able to do much more.
For example, if the manned mission would return with all the data, they would need only storage space, not a high bandwidth connection to Earth to
send back the results, but the restrictions on the temperature and radiation (for example) imposed on the cameras would be more or less the same.
And name for me ``specific jobs that robots could do better`` I`d love to see some examples.
You have RiSE, that can climb vertical walls.
Or the
Inchworm Deep Drilling System, developed (or being developed) by the same company that
made the Rock Abrasion Tool (RAT) for the rovers.
They can also reach places we cannot (like the rescue robots) and endure conditions we can not (extreme heat, pressure, explosions).
And we are much much more agile than machines, here`s an example from wiki regarding the Spirit rover. ``As of sol 1736 (November 20, 2008),
Spirit's total odometry was 7,529 metres (4.68 mi). Wow, big numbers, it took the rover 1736 days to cover 5 MILES, that`s nothing! A human crew
could cover that in one day on some type of rover vehicle.
If it's "on some type of rover vehicle" then it's not a human characteristic,
the physical advantage is from the "rover vehicle", not from the fact that was a human on board.
I didn`t go out of my way to mention that because I though it was obvious, we could tend to each other and take care of the injury on the ship
or base camp, robots obviously couldn`t fix each other.
Robots can fix each other, they just have to have that capability included in them,
there is nothing against that. Some robots can fix themselves and/or build copies of themselves (but for that they would need access to the parts).
Do you mean that a person on a rover can cover longer distances than an robotic rover?
Yes, I though that was quite clear. It
was not clear enough to me, English is not my native language and sometimes I have doubts about what other people mean.
As I said before, if you take a rover that was supposed to be manned and turn it into a robot (like those cars in the DARPA grand challenge) there is
nothing that forces them to have a smaller range and/or speed, unless the electronics bring some limitations that the presence of a human being would
not bring, but considering the DARPA challenges I think that is not a problem, even today and with limited resources like those teams had.
Hope that helps! 
