Re:NASA considers manned outpost around moon to control robo...
by rbanffy on Monday February 13, 2012 @09:58AM
(#3585776)
If I remember correctly, you can put the repeater on orbit around L2 and have it visible from Earth(it would look like it's circling the Moon). Latency may prove a bitch.
But first problems first - shouldn't we put a couple rovers on the near side of the Moon first to develop the technology before we think about (and spend money on) putting them on the other side and invent a couple other problems we really don't have yet (and may never actually have)? Just imagine rovers get brainy enough to be completely autonomous (something I classify as highly likely) - why would we need to have a couple humans on L2 to remote control them?
Re:NASA considers manned outpost around moon to control robo...
by maaku on Monday February 13, 2012 @11:00AM
(#3586037)
On the order of seconds, which is really not a problem (I work at NASA and we've tested remote vehicle operation in analogue environments with 3+ second RTT. The operators adjust to it pretty fast).
Re:NASA considers manned outpost around moon to control robo...
by m3koval on Monday February 13, 2012 @11:44AM
(#3586214)
Nothing is cheap about autonomy, especially when there is limited computing power on-board the robot and power consumption is a major constraint.
Using a hybrid remote operated/autonomous approach is reasonable, though: only send high level commands to the robot. These commands require less bandwidth and are less impacted by latency, so it's the best of both worlds. I believe NASA already does something like this on the Mars rovers by allowing the operator to specify waypoints and letting the robot plan their own paths between the waypoints.
How about we just put a repeater at L4 and L2 and leave the humans home? Or three repeaters in lunar orbit if we don't want to clutter up L4 and L2.
But first problems first - shouldn't we put a couple rovers on the near side of the Moon first to develop the technology before we think about (and spend money on) putting them on the other side and invent a couple other problems we really don't have yet (and may never actually have)? Just imagine rovers get brainy enough to be completely autonomous (something I classify as highly likely) - why would we need to have a couple humans on L2 to remote control them?
"Because it's cool" doesn't cut it.
Using a hybrid remote operated/autonomous approach is reasonable, though: only send high level commands to the robot. These commands require less bandwidth and are less impacted by latency, so it's the best of both worlds. I believe NASA already does something like this on the Mars rovers by allowing the operator to specify waypoints and letting the robot plan their own paths between the waypoints.