A sad thought but perhaps a nessisary one

>> > You may not need many people in orbit anyway. Teleoperated
>robots
>> > could do most of the work controlled from the ground. You
>don't
>> > need to feed a robot.
>>
>AI research still has a way to go before it can replicate the
>decision-making ability of a human being. For teleoperated robots,
>there is the time lag and communications windows to consider; not
>impossible to control a robot that way, but very difficult.
>
>If you see that a robot is about to make a move that will damage it,
>then teleoperation can help as long as the time lag is less than the
>time required for the robot to take that action; in the case of a
>rover on Mars or an asteroid though, with a time lag measured in
>minutes, your robot's toast.

What situation could you think of, were a robot becomes toast on an
asteroid? The American Ion thruster powered probe NEAR did land on the
asteroid EROS even it was not engineered to do a landing.

I think teleoperation would be possible, using ion thruster powered
robots on asteroids. The Ion thruster of SMART I for instance produces
about 75 milli Newton (7,5 grams).
So if you "crash" with such a tender power onto an asteroid, what
should happen?

The robot has to be built that way that he can take a bit.

It is like driving a fairground scooter, not even since the forces
involved are a lot weaker. And yet those forces are enough to get tons
of asteroid materials off the asteroid ground.

The japanes space probe Muses C is ion thruster powered with 4 ion
thrusters. It will try to take samples and we will see next year if
this works. If it works why shouldnt taking some more samples as in
asteroid mining work also.

So tell me a real situation were an robot on an asteroid needs to act
in a hurry.

Frank