NASA Future In-Space Operations: Telerobotics from Mars Orbit
NASA FISO Presentation: Telerobotics from Mars Orbit – Lessons from Robotic Exploration for Human Missions to Mars
“Now available is the August 31, 2016 NASA Future In-Space Operations (FISO) telecon material. The speaker was Jim Bell (ASU) who discussed “Telerobotics from Mars Orbit – Lessons from Robotic Exploration for Human Missions to Mars“.
Note: The audio file and presentation are online and available to download.
“the unfortunate truth is that most things our rovers can do in a perfect sol a human explorer could do in less than a minute” Steve Squires
Dr. Squires’ comment was certainly true at the time (2005), but I suspect he was as much expressing frustration with the available rover technology as suggesting that it would be nice to go there in person. Keep in mind that we have yet to launch a rover with the intelligence of a modern cell phone.
We need to be objective. Uber will be running driverless cabs in a couple of years on public streets. The only factor that limits autonomous mobility by rovers is computing power, and giving Mars rovers the intelligence we already have in Earth rovers might make more sense than sending humans all the way to Mars orbit to control them.
I’ve followed your thinking on this subject over the past few years but I wonder:
Are self-driving cars a good analogy? My wife’s Tesla seems like a marvel when on autopilot, but the underlying technology isn’t very complicated, or new; the underlying sensors and equipment, including radar, are old tech. The car follows the painted lines on the street (and can have trouble on the expressway when the lines break for an exit ramp, and no idea how it would perform in snowpack), looks for known sign shapes with known shape texts, and watches out for heavy metal objects in its vicinity. It ‘seems’ to read complex high-diverging signs overhead, showing them on the dash while highlighting the proper route; but all it’s doing is recalling images from a data base based on the car’s location.
Cool, yes, mostly cool because the application is new. But not difficult particularly; self-driving tech like Uber is a step above, certainly, but again it relies on known street maps, albeit in 3D, while paying particular attention to anything moving. A step, but hardly a conceptual leap.
On the other hand, I wonder if my cellphone could, say, differentiate unknown rock layers aside from catalogue building? Could it see and explore a geologic structure that could contain life? Can it assess the unknown, beyond comparing to – what? Would it add 2+2- without a priori programming? And how would UberMars navigate into the unknown without GPSMars and without road pavement or maps?
Wow, you’ve got a Tesla? What does it do if the car ahead slows down? Or there is a cow in the road?
Seriously, let’s look at some plausible strategies.
1. Long range navigation and planning – https://www.google.com/mars/ is already available and the region near the rover can be mapped with considerable precision from orbit.
2. Position determination: With multiple orbiters with flexible communications systems, position finding at frequent intervals is possible. Landmark sightings can be used when satellites are not visible.
3. Short range navigation: In the clear air of Mars, laser radar can plot the surrounding terrain in three dimensions
4. Objects of interest; multispectral imaging can identify anamolous regions which can be targeted by spectroscopic laser microprobe analysis.
5. Delayed guidance from Earth: If an object of interest is spotted in the imaging sent back to Earth, the rover can store its movement information and quickly retrace its steps.
Tesla: It automatically stays up with traffic through situational awareness; it knows the roadway, and it knows about adjacent vehicles, all reported on the dash; the radar sees the car ahead, of course, but also sees under the first car in front, sensing the second car in front in order to better anticipate.
You can adjust how close you’d like the car get to a leading vehicle; it’s 4 lengths from the factory but 6 is better. It sees mass, too, I should have said, aided by contrast with the background (which is partly how that accident happened; the broadside of the truck was in front of a low and very bright sun). The lane-changing trick is particularly cool. And importantly: Tesla is collecting data from every Tesla on the road via embedded internet via cell tower and at no owner expense, which allows you to use the panel as a browser when underway.
Whew. Back to space:
I know that Curiosity has been the test bed for some autonomous driving but don’t know how it is working. The main thing with Mars: nobody has ever seen it from ground level and high granularity.
1. Considerable precision? GPS is nominally accurate to about 3 meters. But iPhone 5 and 6 use A-GPS: cell towers pass coordinates; plus these phones also use GLONAS. And location on Terra is aided by cues not present on Mars. A very expensive infrastructure, meaning that…
2. …multiple orbiters would approach GPS, I guess, but the number would be quite large if earth is a guide;
3. Short-range: probably true. But don’t forget: a car on Terra can assume that the road grade won’t change much and will never, ever, unexpectedly end at the end of a cliff!
4. Spectra would be the main “differentiator”, I guess; you’re the scientist, not me. But how does a computer recognize the unrecognizable?
5. Earth would have to review everything it did. I suppose an autonomous vehicle could eliminate the obviously unrecognizable, though.
Compared to Mars, Tesla has it easy.
Opportunity has been on Mars for 12.65 years, and it has a total odometry (as of September 13th) of 26.92 miles (43.33 kilometers).
An average person could easily walk that in 3 days. A geologist could probably follow Opportunity’s route and make all the same discoveries it did in a week or two.
There’s a complex set of opportunity costs invovled with sending people to Mars, it will be tremendously more expensive, but we will learn a tremendous amount, and at a far greater rate than we ever could with a rover, even if that rover can drive autonomously.
It’s worth mentioning that both Opportunity and Curiosity have driven in autonomous mode (Opportunity had the benefit of several very long, mostly straight drives, first to Victoria Crater and then to Endurance Crater, where the team used autonomous driving mode for weeks on end), and the Curiosity team has been experimenting with software that lets the rover do some basic on-board image processing, then use that data to select targets of interest and zap them with the laser without being commanded from Earth to do so. We’re getting there.
It seems doubtful that anyone could walk over 40 kilometers on Mars in a spacesuit in three days. It would require a vehicle, and the vehicle would need a navigation system. But would the vehicle really need a driver?
The hardest thing Curiosity did was the landing, and that was autonomous. No human could have landed the Falcon booster stages.
I don’t claim that one solution fits every problem. But an autonomous rover would be a step toward feasible exploration of every solar system object with significant gravity and a solid surface, most of which will be inaccessible to humans for a long time to come.
An average person can easily walk 15 kilometers in a day. A fit and well trained person can walk a marathon in about 7 hours, which is 42 kilometers. Wearing a spacesuit should be no problem for a fit and well trained astronaut, if they could not walk 15 kilometers in a day while wearing a spacesuit I’d even venture to say something might be wrong.
A truly autonomous rover would definitely speed up exploration. Rover operations teams could spend their time selecting targets for study, rather than planning drive routes, for example. We have taken some steps in that direction, but there’s a very long way to go yet.
The current EVA suits and life support (with eight hours of consumables) have a mass around 120 kg. The Apollo suits (six and half hours of consumables, if memory serves) were about 90 kg. Even on Mars, that’s the equivalent weight of 35-45 kg on Earth (75-100 pounds.) That does not include any additional equipment for doing useful work (e.g. field geology) along the way.
An average person can not hike 15 km (a hair under 10 miles) per day, over completely unprepared terrain (which is very different from hiking along a trail), while carrying over 75 pounds of gear. A fairly healthy person probably could. Also, I suspect anyone who can walk a marathon in seven hours is doing so on a paved road.
Also, the comment was about walking 40 km on Mars in three days, not 15 km in one day. Presumably that would involve carrying supplies for all three days as well as some gear to camp out overnight. That is, unless we’re talking about a system of caches and huts along the way.
The Apollo 12 crew covered a total of about 2km on foot in two days of EVAs. In the last of six Apollo lunar landings, the Apollo 17 astronauts, using a vehicle and outside for between 7 and 8 hours each day, traversed a total of 35km in 3 days. http://www.lpi.usra.edu/lun…
Now granted that’s almost as far as Opportunity has gone in a dozen years on Mars (46km) but the astronauts rode most of the way and stopped to make observations at about two dozen locations.
Keep in mind the rovers were sent to areas with more or less flat terrain. Curiosity has started venturing into a pretty rugged area but that’s only been in the past calendar year.
Likely gear for any astronaut on Mars will be a rover. This hasn’t been about whatever other gear they have, it’s whether or not walking that distance is possible. It clearly is.
I think it’s important to do better telerobotics, but I don’t think you can justify as Mars orbital telerobotics crewed mission unless it’s using hardware that can be used for a range of crewed missions (like true reusable interplanetary spaceships).
Otherwise, it’s stuck in the middle. Not as versatile and capable as a crewed landing mission, not as affordable as as remotely-controlled-from-Earth robotics mission. Especially since we could heavily improve our abilities in the latter with better relays and a change in how we control rovers (JPL had an excellent presentation on this, drawing from video games servers’ experience). In fact, the case for a human exploration is only going to get weaker over time, unless we get drastic reductions in the cost of sending people into space in general and Mars in particular (something Paul451 usually likes to remind us of here in the comments).
Maybe.
But looking through the eyes of a rover does little to quell the yearning of a human heart.
It definitely doesn’t have the charm and adventure of a crewed mission, and I’d love to send one to Mars and elsewhere.
Watched a movie called ‘Report from Europa’ last night. I’m a sucker for these just to see the tech, although zero-lag radio to Earth was a cool tech 🙂
The whole idea of a NASA Mars mission anytime before 2050 is a delusion. With only 3 Orion’s funded in the next decade how would you certify the spacecraft to go more than a couple weeks, more than lunar distance? In actuality it was never designed for Mars. And why would anyone believe it would make sense to send people to the vicinity of Mars to operate rovers on the surface that could as well be fully automated or driven from Earth. The whole idea is ludicrous. But, with no real HSF program left anymore, I guess people are free to fantasize. What a waste.
This is another “make work” mission for SLS/Orion because we’re too cheap to build a lander.