Help - Search - Members - Calendar
Full Version: "Dragonfly" Titan explorer drone
Unmanned Spaceflight.com > Outer Solar System > Saturn
Jaro_in_Montreal
Is there a specific website for this Johns Hopkins University Applied Physics Laboratory (APL) concept for a Titan explorer drone?
Looks to be an RTG powered machine, somewhat reminiscent of MSL Curiosity (RTG sticking out the tail end).
But no camera mast, ChemCam, or sampling arm visible in the concept illustration.

QUOTE
Dec. 20, 2017
RELEASE 17-101
NASA Invests in Concept Development for Missions to Comet, Saturn Moon Titan
Dragonfly
Dragonfly is a drone-like rotorcraft that would explore the prebiotic chemistry and habitability of dozens of sites on Saturnís moon Titan, an ocean world in our solar system.
Elizabeth Turtle from the Johns Hopkins University Applied Physics Laboratory (APL) in Laurel, Maryland, is the lead investigator, with APL providing project management.


https://www.nasa.gov/press-release/nasa-inv...turn-moon-titan

Click to view attachment
elakdawalla
dragonfly.jhuapl.edu
rlorenz
QUOTE (elakdawalla @ Dec 20 2017, 04:33 PM) *


Note especially the quite detailed article that went online there this morning.(jump/scroll to resources)
Julius
This mission should easily top the list. Titan here we come! cool.gif
Y Bar Ranch
I use Titan as a case study for an aero class I teach, and am super-psyched at the idea of such a probe. Low gravity and high density are a rotorcraft's best friends.

Already drooling over the detailed 3D photogrammetry extracted from aerial images.
vjkane
QUOTE (Julius @ Dec 22 2017, 07:55 AM) *
This mission should easily top the list. Titan here we come! cool.gif

The science for a comet sample return is very compelling (as is the science for Dragonfly; it comes down to do you prefer a great apple or a great banana?).

And I'd never bet on an easy competition with any proposal lead by Squyres, and he's devoted much of his time the last three years putting his comet sample return proposal together.
Julius
QUOTE (vjkane @ Dec 30 2017, 06:39 AM) *
The science for a comet sample return is very compelling (as is the science for Dragonfly; it comes down to do you prefer a great apple or a great banana?).

And I'd never bet on an easy competition with any proposal lead by Squyres, and he's devoted much of his time the last three years putting his comet sample return proposal together.
. No disrespect to Squyres, but I can already imagine drone flying over titan lakes and magic Island plus extra miles of vistas to image and investigate. This is too good to let go and yes to me is definitely more compelling than the comet sampling mission.
mcaplinger
I share everyone's excitement, but we are unlikely to get a lot more public information about either of these missions before the downselect, and the decision isn't made based on popularity. You can go back historically and look at which missions were competing and which were selected, but even if there are clear patterns there, that's not a great indication of future decisions.
Jaro_in_Montreal
QUOTE (Julius @ Dec 30 2017, 12:23 PM) *
I can already imagine drone flying over titan lakes and magic Island plus extra miles of vistas to image and investigate.


From the description of Dragonfly in http://dragonfly.jhuapl.edu/docs/DragonflyTechDigestAPL.pdf it would not be able to go anywhere near the polar lakes region, landing instead in the equatorial dune fields.

QUOTE
Although the exploration of Titanís seas had previously been considered, notably by the APL-led Titan Mare Explorer (TiME) Discovery concept, the timing mandated by the announcement of opportunity precluded such a mission.
Specifically, with launch specified prior to the end of 2025, Titan arrival would be in the mid-2030s, during northern winter.
This means the seas, near Titanís north pole, are in darkness and direct-to-Earth (DTE) communication is impossible.


Like TIME, Dragonfly also proposes direct-to-Earth (DTE) communication.

Maybe a south-polar visit might be feasible ? ....Ontario Lacus ??

Click to view attachment
mcaplinger
QUOTE (Jaro_in_Montreal @ Dec 30 2017, 09:30 AM) *
Maybe a south-polar visit might be feasible ? ....Ontario Lacus ??

From the article:
QUOTE
Arrival at Titan in the mid-2030s with DTE communication suggests a low-latitude landing site. This
requirement means a similar location and season to the Huygens descent in 2005, so the wind profile and
turbulence characteristics measured by the Huygens probe are directly relevant. Furthermore, the sand
seas that girdle Titanís equator are both scientifically attractive and favorable in terms of terrain characteristics for landing safetyóindeed, it was for these reasons that the 2007 Flagship Study identified these dune fields as the preferred initial target landing area.


And it's unlikely that the vehicle will have enough range to fly from equator to pole.
vjkane
QUOTE (mcaplinger @ Dec 30 2017, 10:03 AM) *
And it's unlikely that the vehicle will have enough range to fly from equator to pole.

I'm not so sure. Assume that Dragonfly lands exactly at the equator. The north pole (and the lake region begins before this) is 4044 km away. Assume that Dragonfly has had a great prime mission and the team is willing to just push it to go the distance. At 40 km per hop once every Titan day (~16 Earth days), the north pole is 4.4 years away.

Given that Titan is pretty benign, once you've solved the problem of how to stay warm (always take your warm MMRTG with you when you visit), the limiting factor on the mission may well be how long the the MMRTG power lasts given radioactive decay.

This map in this conference abstract suggests targets that might be in the range of a primary or a first extended mission.

LPSC 2017 abstract

Explorer1
Opportunity certainly went beyond its designed range, I wouldn't be surprised that a flying vehicle could go so much farther.
I would be more worried that perhaps the environment isn't so benign; Oppy dealt with dust storms, but what about possible flash floods (as the Huygens landing site showed)? Without weather observations from orbit, that would be a nasty surprise outside the equatorial dune seas! Or the rotors' reaction to giant raindrops in-flight....
nprev
I wonder how effectively it could navigate over long distances. We don't have nearly good enough surface maps for AI terrain recognition, there's no significant magnetic field, so all that's left is inertial. Maintaining a good heading alignment over long periods may be problematic since IMUs do have inherent drift, and though periodic realignment is the usual method to correct that Titan's outer shell rotation seems to vary significantly in comparison to the rest of the moon's mass (not sure if that's a fixed offset or variable), and measuring rate & direction of rotation after vertical alignment is the usual method of finding true north (and latitude).

This could possibly be augmented by RDFing the vehicle's downlink to Earth, but not sure how much position precision could be achieved...tens/hundreds of km? Then again, maybe the position of the Sun could be used as well, foggy though it's gonna be. Dunno if Saturn would be detectable, but the Sun's definitely gonna be the only possible reference star.
HSchirmer
QUOTE (nprev @ Dec 31 2017, 09:31 AM) *
We don't have nearly good enough surface maps for AI terrain recognition, there's no significant magnetic field, so all that's left is inertial.


Not necessarily, just old-school triangulation: "1800s mountain peak GPS", using trigonometry to track where the mountain peaks are on the horizon.

Ala "the Englishman who went up a hill, but came down a mountain" you build up a triangular grid of the highest points by surveying.
That lets you triangulate your map location, and calculate your height, based where they are on the horizon.
RoverDriver
QUOTE (nprev @ Dec 31 2017, 01:31 AM) *
...
but the Sun's definitely gonna be the only possible reference star.


The nadir vector can be detected by the accelerometers. The Mars rovers we use: clock, Sun position, and nadir vector. As an alternative gyro compassing might be quite more difficult but not impossible.

Paolo
vjkane
QUOTE (nprev @ Dec 31 2017, 01:31 AM) *
I wonder how effectively it could navigate over long distances. We don't have nearly good enough surface maps for AI terrain recognition, there's no significant magnetic field, so all that's left is inertial. Maintaining a good heading alignment over long periods may be problematic since IMUs do have inherent drift, and though periodic realignment is the usual method to correct that Titan's outer shell rotation seems to vary significantly in comparison to the rest of the moon's mass (not sure if that's a fixed offset or variable), and measuring rate & direction of rotation after vertical alignment is the usual method of finding true north (and latitude).

This could possibly be augmented by RDFing the vehicle's downlink to Earth, but not sure how much position precision could be achieved...tens/hundreds of km? Then again, maybe the position of the Sun could be used as well, foggy though it's gonna be. Dunno if Saturn would be detectable, but the Sun's definitely gonna be the only possible reference star.

From Ralph et al.'s paper, Dragonfly would do 40 km hops with 16 days between. I presume the quadcopter would have its position updated during the between days.

The paper hints that longer flights are likely possible and 40 km is the safe planning distance. One factor that would shorten traverses is the plan to use each flight to locate a more distant future landing site and then fly back to a previously scouted nearer landing site. With experience, the mission team might gain the confidence to not do the fly back and allow the quadcopter to chose its own safe landing site. With lidar or structure from motion (building 3D maps from stereo images), Dragonfly could continuously search for safe landing sites below its flight path and know of safe landing sites.


mcaplinger
QUOTE (RoverDriver @ Dec 31 2017, 08:25 AM) *
The Mars rovers we use: clock, Sun position, and nadir vector.

That gives you rover orientation for antenna pointing, but AFAIK, not absolute location to any kind of accuracy.

For Titan, I would expect Earth-based radiometric positioning to be accurate to at least 100s of meters, easily good enough for vehicle navigation.
HSchirmer
QUOTE (vjkane @ Dec 31 2017, 04:46 PM) *
From Ralph et al.'s paper, Dragonfly would do 40 km hops with 16 days between. I presume the quadcopter would have its position updated during the between days.

The paper hints that longer flights are likely possible and 40 km is the safe planning distance. One factor that would shorten traverses is the plan to use each flight to locate a more distant future landing site and then fly back to a previously scouted nearer landing site. With experience, the mission team might gain the confidence to not do the fly back and allow the quadcopter to chose its own safe landing site. With lidar or structure from motion (building 3D maps from stereo images), Dragonfly could continuously search for safe landing sites below its flight path and know of safe landing sites.


Well, when it comes to auto-navigation, you really have to check out U-Penn's GRASP program, and the Kumar lab's drones...
https://www.grasp.upenn.edu/research-groups/kumar-lab
They've done some really neat work, check "Journal of Field Robotics"...
And their youtube channel

https://www.youtube.com/watch?time_continue...p;v=rJfQncmWpCo

IIRC, somebody had the brilliant idea to modulate the prop speed among the 4 blades to generates a beat tone
for sonar range finding. The drone "listens" for the echo to measure distance to large objects.

Nice coincidence that Earth and Titan have nitrogen atmospheres, acoustics shouldn't be that different...
nprev
QUOTE (vjkane @ Dec 31 2017, 08:46 AM) *
With experience, the mission team might gain the confidence to not do the fly back and allow the quadcopter to chose its own safe landing site. With lidar or structure from motion (building 3D maps from stereo images), Dragonfly could continuously search for safe landing sites below its flight path and know of safe landing sites.


Interesting, and thanks for the responses, all. Didn't know that terrestrial-based radiometry was accurate at sub-km resolution, Mike, so that solves the main problem: navigating to targets like lakes and cryovolcano candidates that may be extremely distant from the original landing site. Periodic position fixes combined with the local-scale 'hop & look' nav methods described should solve that with a high degree of precision and operational safety.
Y Bar Ranch
Wonder what kinds of information can be gathered by just going into a low hover or running the rotors on the ground to generate some airflow. Properties of surface particles? Etc?
vjkane
QUOTE (Y Bar Ranch @ Dec 31 2017, 08:39 PM) *
Wonder what kinds of information can be gathered by just going into a low hover or running the rotors on the ground to generate some airflow. Properties of surface particles? Etc?

That is specifically mentioned in the paper Ralph gave the link to a few posts up.
Y Bar Ranch
QUOTE (vjkane @ Jan 1 2018, 10:30 AM) *
That is specifically mentioned in the paper Ralph gave the link to a few posts up.

Ahhh, missed it on the first read.
Daniele_bianchino_Italy
I certainly hope for this mission. but ... if one of the most extraordinary things in the solar system are the Titan lakes and seas, why spend it on a mission for dry Titan areas ?
I do not really understand, 99% of us are hoping to see lakes closely. Why after many years of waiting do a mission on Titan in area without lakes? ... bha!
RoverDriver
QUOTE (mcaplinger @ Dec 31 2017, 08:50 AM) *
That gives you rover orientation for antenna pointing, but AFAIK, not absolute location to any kind of accuracy.
...


True, although maybe you can get latitude, definitely not longitude. Likely dead reckoning would be quite difficult unless some kind of visual odometry or SLAM is employed.

Paolo
fredk
QUOTE (RoverDriver @ Jan 2 2018, 07:23 PM) *
definitely not longitude

Knowing the time, couldn't you also get the longitude? We'd need the sun's elevation (I guess from imaging in some IR band, if possible) relative to the nadir (from accelerometers) for a few observations. Of course the precision won't be good - one degree relative precision of the sun's position translates to about 45 km position accuracy on the surface, so it sounds like the radio approach would be more precise.
mcaplinger
QUOTE (fredk @ Jan 2 2018, 10:54 AM) *
Knowing the time, couldn't you also get the longitude?

Certainly (see https://en.wikipedia.org/wiki/Longitude_(book) ) but as I noted this is not to any accuracy and AFAIK has never been used for Mars rover positioning as there are better ways to do it.

I'm not sure you can position the sun very accurately with imaging on Titan, but my point is, you don't have to.
Explorer1
QUOTE (Daniele_bianchino_Italy @ Jan 2 2018, 12:51 PM) *
I certainly hope for this mission. but ... if one of the most extraordinary things in the solar system are the Titan lakes and seas, why spend it on a mission for dry Titan areas ?
I do not really understand, 99% of us are hoping to see lakes closely. Why after many years of waiting do a mission on Titan in area without lakes? ... bha!

I believe the issue is the seasons; Saturn (and Titan) will enter northern winter by the time the mission arrives, which not only means it is dark, which makes it tougher to run a mission without extra lights, but there is also no direct line to communicate with Earth (without a relay satellite, which would be quite expensive). The last chance this Saturnian year was Titan Mare Explorer, but it was obviously not selected in the last round of Discovery proposals, so it will be a wait until the northern lakes are illuminated again.

Ontario Lacus is in the southern hemisphere, but it is much smaller and shallower than its northern counterparts. Other more equatorial lakes have been theorized but not yet confirmed
vjkane
QUOTE (Explorer1 @ Jan 2 2018, 12:49 PM) *
I believe the issue is the seasons; Saturn (and Titan) will enter northern winter by the time the mission arrives, which not only means it is dark, which makes it tougher to run a mission without extra lights, but there is also no direct line to communicate with Earth (without a relay satellite, which would be quite expensive). The last chance this Saturnian year was Titan Mare Explorer, but it was obviously not selected in the last round of Discovery proposals, so it will be a wait until the northern lakes are illuminated again.

Ontario Lacus is in the southern hemisphere, but it is much smaller and shallower than its northern counterparts. Other more equatorial lakes have been theorized but not yet confirmed

How far south does the most southern northern lake go? Would that be outside the polar night? (Sorry, don't have time to go look at a map and compare to the axial tilt.)
Webscientist
QUOTE (vjkane @ Jan 2 2018, 11:41 PM) *
How far south does the most southern northern lake go? Would that be outside the polar night? (Sorry, don't have time to go look at a map and compare to the axial tilt.)


I had the same question in mind.

I've taken a look at a map of 2016.
It seems that Kraken Mare has extensions at about 60 degrees north latitude, roughly the equivalent to the top of Scotland (Ralph must know).
But the axial tilt of Titan is a bit higher than that of the Earth (27 degrees versus 23.4 degrees).
So wha



Webscientist
QUOTE (Webscientist @ Jan 3 2018, 07:05 PM) *
I had the same question in mind.

I've taken a look at a map of 2016.
It seems that Kraken Mare has extensions at about 60 degrees north latitude, roughly the equivalent to the top of Scotland (Ralph must know).
But the axial tilt of Titan is a bit higher than that of the Earth (27 degrees versus 23.4 degrees).
So for the next good exploration window, maybe in the 40s.
But if there is the will...
2017-2024 was the perfect time I guess.

Daniele_bianchino_Italy
iL kraken sea extends to 56 north. I think small lakes can be on 50 north.
scalbers
How about some twilight lake watching?

https://www.space.com/36609-twilight-outshi...moon-titan.html

Although the paper mentioned in this article is mainly referring to the total disk brightness as seen from space, it is a reminder that a reasonable amount of scattered light is available at the surface during twilight (e.g. in near-polar winter).
Habukaz
QUOTE (Explorer1 @ Jan 2 2018, 09:49 PM) *
Ontario Lacus is in the southern hemisphere, but it is much smaller and shallower than its northern counterparts. Other more equatorial lakes have been theorized but not yet confirmed


Polaznik Macula, Sionascaig Lacus and Urmia Lacus on Google Maps

This mission seems like an excellent opportunity to test the lake hypotheses for these two features. wink.gif

Also of note:

QUOTE
However, Stofan et al. (2007) and Tan et al. (2013) state that liquid methane is thermodynamically stable anywhere on the surface of Titan.


So, maybe there could be smaller pools of liquid, or even smaller lakes too small to have been resolved yet, even closer to the equator.



JRehling
From half of Titan's surface, Saturn will be visible almost all the time. That seems like it'd be very useful for navigation, even more so than the Sun, because the Sun will vanish for ~192 hours at a time.

Going a lot farther down the magnitude scale, a really interesting possibility would be if you could see Betelgeuse, Antares, Aldebaran, and possibly some other stars like Arcturus. The former are red giants that are bright in infrared, which, as we know, penetrates Titan's haze pretty well. You'd never see them in the daytime sky, but at night they'd be brighter in the IR band than they are in visible light from Earth. Seems like navigating by the stars could cover your nights on Titan and the combination of the Sun and Saturn would handle the daytime. And an IR sensor could be pretty sensitive operating at 94K.

In addition, the radio link with Earth would give you greater precision longitude checks twice per sol.
vjkane
QUOTE (JRehling @ Jan 6 2018, 12:15 AM) *
Going a lot farther down the magnitude scale, a really interesting possibility would be if you could see Betelgeuse, Antares, Aldebaran, and possibly some other stars like Arcturus. The former are red giants that are bright in infrared, which, as we know, penetrates Titan's haze pretty well. You'd never see them in the daytime sky, but at night they'd be brighter in the IR band than they are in visible light from Earth. Seems like navigating by the stars could cover your nights on Titan and the combination of the Sun and Saturn would handle the daytime. And an IR sensor could be pretty sensitive operating at 94K.

The descriptions so far don't mention an IR sensor, although it could be considered an engineering instrument. The highest frequency atmospheric window is 0.93 microns. The Mastcam Z cameras goes to 0.88 microns. I don't know if the sensor itself goes to 0.93 microns and the Mastcam Z limit is based on the scientific value of that band and not the sensor.
vjkane
QUOTE (vjkane @ Jan 6 2018, 10:26 AM) *
The descriptions so far don't mention an IR sensor, although it could be considered an engineering instrument. The highest frequency atmospheric window is 0.93 microns. The Mastcam Z cameras goes to 0.88 microns. I don't know if the sensor itself goes to 0.93 microns and the Mastcam Z limit is based on the scientific value of that band and not the sensor.

I'll correct my last post after finding a better paper. The MastCam Z sensor goes out to a full micron, so if Dragonfly uses a similar sensor, it could image through the 0.93 band. If nothing else, it could take great tourist pictures of the surface with Saturn in the sky (although it would be a monochrome image).
scalbers
At around 0.93 microns these images would be interesting, though still a bit hazy since the aerosol optical depth at this wavelength is about 3 at the zenith. Thus looking at Saturn and stars would be a bit like the view through medium-thin cirrus clouds on Earth. Some details are in figure 12.18 from this paper: http://ciclops.org/media/sp/2010/6514_15623_0.pdf. Saturn may look best during twilight and it should be high in the sky. A wide angle lens (or a mosaic) would help with showing the terrain at the same time.
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.
Invision Power Board © 2001-2018 Invision Power Services, Inc.