In This Issue
Fall Issue of The Bridge on Space Exploration
September 1, 2021 Volume 51 Issue 3
Close collaboration between engineering and science has enabled marvels of space exploration over decades. Eight exemplary missions are described in this issue, conveying the excitement, challenges, and breakthroughs involved in efforts to better understand the wonders and mysteries of this solar system.

The Dragonfly Mission to Titan: Enabled by Scientific Discovery and Engineering Development

Monday, September 13, 2021

Author: Elizabeth P. Turtle and Ralph D. Lorenz

Planning of the Dragonfly mission to Saturn’s moon Titan benefited from decades of scientific observations and technological progress.

“NASA plans amazing thing for Destination X” is a frequent attention-grabbing headline format. Many possible mission concepts are considered for many places in (and beyond) this solar system. But how does a concept make the transition from idea to mission project—with a detailed engineering design, budget, and schedule—and eventually to hardware on another world?

The idea that a robotic rotorcraft (imagined early on as a helicopter) might be the best platform to explore Saturn’s largest moon Titan arose in the 1990s (e.g., Lorenz 2000; Young 2001), stimulated both by contemplation of Titan in preparation for the joint NASA-ESA Cassini-Huygens mission (launched in 1997) and by the engineering development of progressively more capable robotic aerial vehicles for terrestrial applications.

But it was still too early. Titan was largely a mystery (Lorenz 2020). At 5150 km in diameter—larger than the planet Mercury—it was the largest unexplored body in the Solar System.

What Was Learned from Early Exploration?

Titan’s thick, hazy atmosphere—four times denser at the surface than Earth’s atmosphere—hid the surface from scrutiny by the Voyager 1 spacecraft that made a close flyby in 1981. Voyager revealed that the atmosphere was mostly nitrogen with a few percent methane, with a surface temperature of 94 K (−189°C) such that methane might condense on the surface as a liquid; it was even speculated that Titan’s surface could be covered by a global ocean (Lunine et al. 1983).

Telescopic data in the mid-1990s had revealed the surface at scales of only several hundred kilometers. But they showed that some areas were darker than others and that these albedo features persisted, demonstrating that large portions of the surface were solid as opposed to a global surface ocean of methane and ethane.

Turtle and Lorenz fig 1.gif

FIGURE 1 (Left) Cassini Imaging Science Subsystem map of Titan’s surface at 938 nm (Karkoschka et al. 2018). Dark areas near the equator are vast fields of dunes of organic sand, and dark areas near the poles are lakes and seas of liquid methane. Image credit: NASA/JPL-Caltech/University of Arizona. (Right) Huygens Descent Imager/Spectral Radiometer image of Titan’s surface (Karkoschka and Schröder 2016). Image credit: ESA/NASA/JPL/University of Arizona.

In 2005 the Huygens probe parachuted down near Titan’s equator into what turned out to be a streambed, with a methane-damp, soft surface strewn with (water-ice?) cobbles, possibly rounded by tumbling in some past flash flood. Off in the distance were a couple of long streaks that turned out to be dunes of organic sand (figure 1). Even in the tiny area viewed by Huygens, the terrain was diverse, and it took several years for Cassini mapping to reveal the broad distribution and relationships of geological features, including methane seas and lakes largely confined to areas near the poles.

Cassini, designed for a 4-year mission at Saturn (2004–08) with brief flybys past Titan roughly once a month, had its mission extended twice: first to 2010 to cover the Saturn system’s equinox season, and then by another season, ending in September 2017 just after the northern summer solstice (Titan years are 29.5 Earth years long, and the obliquity of the Saturn system is 26.7°).

The fleeting glimpses of Titan during Cassini’s 13 years and 120+ flybys did not permit nearly the complete survey possible with an orbiter around Titan. In some respects they only elevated the question of Titan’s unknown surface composition, compared to other Titan science goals on which Cassini made more headway than had been expected.

What Would Be the Right Type of Mission?

The “right” mission type for Titan was far from certain; figure 2 illustrates a variety of options considered. NASA and ESA commissioned concept studies of large “flagship” missions to Titan (Coustenis et al. 2009; Leary et al. 2008; Reh et al. 2009), Enceladus (­Razzaghi et al. 2007), Europa (Clark et al. 2007, 2009), and ­Jupiter ­System/Ganymede (Kwok et al. 2007), of which the ­Europa ­studies eventually led to the Europa Clipper mission scheduled for launch in 2024 (Pappalardo et al. 2021).

Turtle and Lorenz fig 2.gif
FIGURE 2 (Top left) Montgolfière or hot air balloon. Image credit: NASA/Tibor Balint. (Top right) A Pathfinder/Beagle-2–inspired dune lander concept using airbags for landing. Image credit: James Garry. (Bottom left) Discovery Phase-A (Titan Mare Explorer, TiME) capsule floating on a Titan sea. Image credit: Johns Hopkins Applied Physics Laboratory. (Bottom right) Aerial Vehicle for In-situ and Airborne Titan Reconnaissance (AVIATR) drone concept. Image credit: Mike Malaska, CC-BY-SA2.0. 

The Titan studies, confronting interesting scientific questions at a range of scales and altitudes, proposed a suite of platforms to match: a lander to analyze the surface chemistry and record weather and seismic activity over a long period; an orbiter to make global observations of the surface and atmosphere; and a balloon to bridge the gulf between these scales.

Balloons for Titan, exploiting its dense atmosphere, had been considered since the 1990s. One neat aspect was that a radioisotope power supply—which would be needed to power the instruments and systems since Titan is so far from the Sun that solar power is ­unviable—would also provide “waste” heat, which could be used to generate buoyancy (just as with hot air balloons on Earth).

But a balloon mission architecture always raises concerns about lack of control authority. Even experienced hot air balloon pilots who exhibit a remarkable degree of control, exploiting subtle clues like bird flight or smoke plumes to judge wind patterns, cannot fly upwind. For a Titan mission, there would be little hope that an autonomous system—a necessity because light and radio signals take more than an hour to reach Titan from Earth, making direct control of flight ­impossible—could achieve the capabilities of a human pilot, especially in an unfamiliar environment.

But the selection of another type of mission depended on resolution of a number of other factors.

How Would Power Be Provided?

A vital element in any long-term Titan in situ mission is radioisotope power. The Huygens probe relied on about 50 kg of batteries for its ~4 hours of operation on Titan, but chemical stored energy is not practical for anything lasting more than a fraction of a Titan day/night cycle (which takes 16 Earth days). The dense, cold atmosphere chills external surfaces and demands both thick insulation and internal heating to maintain benign temperatures for equipment.

Radioisotope power systems, using the decay of ­plutonium-238 as a heat source for an energy con­verter, enable long-term missions. However, the thermo­electric converters on Voyager and Cassini (using silicon-­germanium semiconductors) function only in a vacuum, and in the early 2000s there was no ready solution to powering Titan landers.

One promising system, central to the design of the proposed Titan Mare Explorer (TiME; Stofan et al. 2013)—a capsule that would float in a Titan sea and measure aspects of its composition, bathymetry, local weather, and air-sea interactions (studied in Phase A under NASA’s Discovery program in 2011–12)—was the Advanced Stirling Radioisotope Generator (ASRG[1]). This technology was also enabling for the autonomous airplane mission concept Aerial Vehicle for In-situ and Airborne Titan Reconnaissance (AVIATR), which would have studied Titan’s surface and atmosphere, flying continuously for a year while beaming aerial camera views and spectroscopic data back to Earth (Barnes et al. 2012). But development of the ASRG stalled.

Only in 2011, when the first multimission radio­isotope thermoelectric generator (MMRTG) was launched on the Curiosity Mars rover, with converters designed to operate in an atmosphere, was an off-the-shelf power system ready for the Titan surface environment. Like the concept that would become Dragonfly (Lorenz et al. 2018), Curiosity exploits the waste heat from the MMRTG both to keep warm during the cold Martian night and to provide about 100 W of electrical power. Curiosity further showed that the engineering challenge of incorporating an MMRTG in an aeroshell, and rejecting the heat from it during the interplanetary cruise with a pumped fluid loop, was tractable.

Where Should a Lander Go to Answer Key Science Questions?

In the mid-2010s Cassini made some remarkable radar observations of Titan’s northern seas (Le Gall et al. 2016; Mastrogiuseppe et al. 2014): it detected an echo not just from the sea surface but also from the sea floor. Not only did this technique measure the depth of that sea at 160 m (a measurement that was to have been done in situ by sonar on TiME), but the very detectability of a bottom echo revealed that the liquid was exceptionally radar-transparent.

Titan’s seas had been expected to be a mix of methane and ethane with various dissolved organic compounds, raising the question of whether some kind of alternative life chemistry could exist in such a liquid. But the transparency showed that the liquid had to be nearly pure methane. Such a composition, probably resulting from methane distilled as rainfall, indicated that the chemistry measurements to be made by TiME might have proven less complex than hoped.

Titan’s surface is far too cold for liquid water to persist, but Cassini data suggest that, beneath the moon’s 50–150 km thick ice crust, there is an internal global water ocean, and some isolated areas at the surface may have seen transient water environments that resulted from either the eruption of internal water (a ­cryovolcano) or melting (due to crater-forming comet impacts).

Even before Cassini’s arrival, the most interesting places to explore from an astrobiological point of view were expected to be those where liquid water might have interacted with the abundant organic compounds produced in the atmosphere (Neish et al. 2018). The breakup of methane and nitrogen in Titan’s upper atmosphere by radiation and solar ultraviolet light leads to the formation of ethane, acetylene, hydrogen cyanide, and a host of compounds, many of which form Titan’s atmospheric haze layers and eventually accumulate on its surface (e.g., as part of its dune sands).

But astrobiologically speaking the compounds in Titan’s atmosphere are something of a dead end because they do not incorporate oxygen (as do most molecules in living things)—the atmosphere has only traces of ­oxygen in the form of carbon dioxide and ­monoxide. However, when analogues of Titan’s haze are exposed to liquid water in the laboratory, they can form amino acids and ­pyrimidine bases, the building blocks of proteins and some of the molecules that encode information in DNA (Cable et al. 2014; Hörst et al. 2012).

How Could a Lander Best Access Areas of Scientific Interest?

As project start dates moved into the mid-2010s, the prospect of a mission reaching Titan during the icy moon’s northern summer in the early 2020s began to fade. This season had been favorable for exploration of Titan’s high northern latitudes not only in terms of solar illumination but also because direct transmission of data from the surface is possible only when Earth is above the horizon. (With an orbiter to act as a relay, a northern seas mission can be done any time, but at much higher cost.) Attention therefore turned to Titan’s lower latitudes.

Although Cassini’s maps show the locations of target features of interest, the best data have a resolution of no better than 300 m (roughly comparable to Mars imaging prior to the Viking landers in 1976), so there would be no guarantee that a single landing site would be safe to access or offer exposures of chemically interesting deposits (figure 3). As elsewhere, the solution is in situ mobility.

Turtle and Lorenz fig 3.gif
FIGURE 3 Dragonfly’s initial landing site is among the equatorial dunes near the 80 km diameter Selk impact crater, where carbon-rich surface material may have had the opportunity to mix with liquid water impact melt (Neish et al. 2018). Image credit: Johns Hopkins Applied Physics Laboratory.

Ground Mobility

Surface mobility makes it possible to land where it is reasonably likely the terrain is safe, and then to exploit much better information obtained in situ to find the most scientifically valuable spots nearby (Lorenz 2019). This is the approach used by rover missions on Mars, and indeed for seafloor exploration on Earth.

A Titan rover is conceivable, but between sand dunes and damp streambeds, the trafficability of much of the moon’s surface might be doubtful, and even on dry Mars the Spirit rover got irretrievably stuck in 2010.

Aerial Mobility

Titan’s low gravity—one seventh that of Earth (a little less than our Moon’s)—and dense atmosphere make it a uniquely easy place to explore using heavier-than-air flight, as was recognized in the 1990s (Lorenz 2000). For these reasons, when NASA solicited mission concepts in 2016, the idea of a Titan rotorcraft (re)emerged and seemed a demonstrably feasible prospect.

No individual technology can take credit here, although one might generalize “the drone revolution”—the ubiquity of cheap camera-toting quadcopters in gadget stores enhances the perception that the achievement of autonomous flight on another world is a natural extrapolation. The details of the drone revolution are somewhat prosaic—better batteries, better motors with rare-earth magnets, and better speed controllers to push power through the motors, among others. Some consumer drone advances have made components that have long been part of space systems cheaper and ­smaller, such as the gyros and accelerometers that let a drone fly stably.

One can also point to significant engineering progress over the last decade, in NASA’s lunar and Mars landers, in machine vision for navigation, and in sensors such as lidar for autonomous hazard detection and avoidance. These capabilities (which have also been exercised by the Chinese moon landers, and to some extent in consumer drones that optically track targets for movie­making) mean that a Titan lander, whether relocatable or not, is not at the mercy of random encounters with adverse terrain. One can use terrestrial analogue environments (as did planners of the Viking landers) to assert that a landing site will probably be safe, but the specter of an unlucky rock in just the wrong place keeps planetary explorers awake at night. Hazard avoidance technology brings terrain risks down to much more acceptable levels.

What Finally Led to the Dragonfly Mission Concept?

What enabled the transition from a few speculative concept papers to a fully developed mission proposal (Lorenz et al. 2018), with preliminary designs, test plans, budgets, schedules, and significant investment by the Johns Hopkins Applied Physics Laboratory?

Scientific Context, More Data, Effective Power

First, of course, was the scientific context. Cassini’s discoveries at Enceladus and Titan, and interest in Europa, drove a unifying theme of “Ocean Worlds” for the NASA solicitation.[2] The theme focuses on habitability of these distant worlds and encompasses aspects of Titan such as chemical interactions with water and geophysical (e.g., seismological) study of the interior ocean. Compared to other Ocean Worlds, Titan’s thick atmosphere makes it easier to deliver instrumentation directly to its surface.

Second, Cassini-Huygens data and a greatly improved understanding of meteorology (e.g., global circulation models) gave much higher confidence in the Titan environment. The arrangement of Titan’s dunes could be understood in terms of its wind patterns, and similarities of its duneforms to some dunefields on Earth allow terrestrial landscapes to be used as analogues to predict slopes and other terrain characteristics, even though the composition is different. This meant that confident and realistic requirements and constraints could be devised for engineering design.

Finally, the MMRTG is the sine qua non of sustained Titan surface operation. This power system offers unique capability, but has not always been available. A mission to fly half a decade after Perseverance (Curiosity’s successor rover at Mars) fit well with the overall radio­isotope power system program.

Other Enabling Advances

Additional developments that facilitate the Dragonfly design include the radial line slot array antenna (figure 4). Mounted on a two-axis gimbal, it offers design simplicity and an extremely low profile; a conventional parabolic dish and antenna feed would cause much more drag in flight.

Turtle and Lorenz fig 4.gif
FIGURE 4 Illustration of Dragonfly rotorcraft lander on the surface of Titan. The radial line slot array antenna folds down flat on top of the vehicle during powered flight. Image credit: Johns Hopkins Applied Physics Laboratory.

Dragonfly also benefits from augmented reality tools that simplify design, fabrication, and integration of spacecraft components.[3]

Machine vision, a modest advance on systems used in cruise missiles since the 1980s, allows more accurate guidance than would be possible with inertial sensing alone (e.g., gyros). Without the GPS navigation taken for granted in terrestrial applications, a Titan rotorcraft would be limited to hops of only a few hundred meters. Through onboard correlation of images with those ­taken in previous flights, it is possible for Dragonfly to be directed to specific sites of interest several kilometers from the takeoff site.

But no artificial intelligence is planned or needed: scientists on the ground will determine where the interesting places are in terms of lander-relative coordinates, and will command Dragonfly to fly there. The onboard autonomy simply ensures adherence to the designated path.

Even in Titan’s dense atmosphere and low gravity, where in principle a suitably equipped human being could flap wings with her own muscle power and fly, a vehicle needs much more than 100 W of power. A practical vehicle could use a large battery to fly for about half an hour and recharge the battery with MMRTG power over the 8-day Titan night, when it is dark and geometry precludes communication with Earth (Lorenz 2000).

More than 99 percent of the time, Dragonfly, like the Viking landers, will stay on the ground performing occasional imaging, surface sampling, and chemical analyses, as well as ongoing seismic and ­meteorological monitoring. But in one brief flight it will be able to fly several kilometers—conceivably farther than a Mars rover has driven to this point.


The prospect of aerial vistas much like those from a balloon, coupled with a lander whose science value is multiplied by being able to access dozens of locations, proved irresistible. Combined with the largely evolutionary rather than revolutionary technological and scientific developments described above, the opportunity arose to propose Dragonfly. A few years earlier it would have been impossible.

The mission is scheduled to launch in June 2027 and projected to land on Titan in the mid-2030s.


Barnes JW, Lemke L, Foch R, McKay CP, Beyer RA, ­Radebaugh J, Atkinson DH, Lorenz RD, Le Mouélic S, Rodriguez S, and 21 others. 2012. AVIATR—Aerial Vehicle for In-situ and Airborne Titan Reconnaissance: A Titan airplane mission concept. Experimental Astronomy 33(1):55–127.

Cable ML, Hörst SM, He C, Stockton AM, Mora MF, Tolbert MA, Smith MA, Willis PA. 2014. Identification of primary amines in Titan tholins using microchip nonaqueous capillary electrophoresis. Earth and Planetary Science Letters 403:99–107.

Clark K, Greeley R, Pappalardo R, Jones C. 2007. 2007 ­Europa Explorer Mission Study: Final Report. Pasadena CA: Jet Propulsion Laboratory.

Clark K, Magner T, Pappalardo R, Blanc M, Greeley R, Lebreton J-P, Jones C, Sommerer J. 2009. Jupiter Europa Orbiter Mission Study 2008: Final Report—The NASA Element of the Europa Jupiter System Mission (EJSM). Washington: NASA and Paris: ESA.

Coustenis A, Atreya SK, Balint T, Brown RH, Dougherty MK, Ferri F, Fulchignoni M, Dautier D, Gowen RA, Griffith CA, and 146 others. 2009. TandEM: Titan and Enceladus mission. Experimental Astronomy 23:893–946.

Hörst SM, Yelle RV, Buch A, Carrasco N, Cernogora G, ­Dutuit O, Quirico E, Sciamma-O’Brien E, Smith MA, Somogyi Á, and 3 others. 2012. Formation of amino acids and nucleotide bases in a Titan atmosphere simulation experiment. Astrobiology 12(9):809–17.

Karkoschka E, Schröder SE. 2016. The DISR imaging mosaic of Titan’s surface and its dependence on emission angle. Icarus 270:307–25.

Karkoschka E, McEwen A, Perry J, Turtle E. 2018. A ­global mosaic of Titan’s surface albedo using Cassini images. American Astronomical Society, DPS meeting #50, id.216.02.

Kwok J, Prockter L, Senske D, Jones C. 2007. Jupiter ­System Observer Mission Study: Final Report. Washington: NASA.

Leary JC, Strain RD, Lorenz RD, Waite JH. 2008. Titan Explorer Flagship Mission Study. Washington: NASA.

Le Gall A, Malaska MJ, Lorenz RD, Janssen MA, Tokano T, Hayes AG, Mastrogiuseppe M, Lunine JI, Veyssière G, ­Encrenaz P, and 1 other. 2016. Composition, seasonal change, and bathymetry of Ligeia Mare, Titan, derived from its microwave thermal emission. JGR Planets 121(2):233–51.

Lorenz RD. 2000. Post-Cassini exploration of Titan: Science rationale and mission concepts. Journal of the British Interplanetary Society 53:218–34.

Lorenz RD. 2019. Calculating risk and payoff in planetary exploration and life detection missions. Advances in Space Research 64(4):944–56.

Lorenz RD. 2020. Saturn’s Moon Titan Owners’ Workshop Manual. Yeovil UK: Haynes.

Lorenz RD, Turtle EP, Barnes JW, Trainer MG, Adams DS, Hibbard KE, Sheldon CZ, Zacny K, Peplowski PN, ­Lawrence DJ, and 8 others. 2018. Dragonfly: A rotorcraft lander concept for scientific exploration at Titan. Johns Hopkins Technical Digest 34(3):374–87.

Lunine JI, Stevenson DJ, Yung YL. 1983. Ethane ocean on Titan. Science 222(4629):1229–30.

Mastrogiuseppe M, Poggiali V, Hayes A, Lorenz R, Lunine J, Picardi G, Seu R, Flamini E, Mitri G, Notarnicola C, and 2 others. 2014. The bathymetry of a Titan sea. Geophysical Research Letters 41(5):1432–37.

Neish CD, Lorenz RD, Turtle EP, Barnes JW, Trainer MG, Stiles B, Kirk R, Hibbitts CA, Malaska MJ. 2018. Strategies for detecting biological molecules on Titan. Astrobiology 18(5):571–85.

Pappalardo RT, Becker T, Blaney D, Blankenship D, Burch J, Christensen P, Craft K, Daubar I, Gudipati M, Hayes A, and 19 others. 2021. The Europa Clipper Mission: Understanding icy world habitability and blazing a path for future exploration. Paper #255 submitted to the Planetary Science and Astrobiology Decadal Survey 2023-2032.

Razzaghi AI, Simon-Miller AA, Di Pietro DA, Spencer JR. 2007. Enceladus: Saturn’s Active Ice Moon—Flagship ­Mission Concept Study. Washington: NASA. 

Reh K, Erd C, Matson D, Coustenis A, Lunine J, Lebreton J-P. 2009. TSSM: Titan Saturn System Mission—NASA/ESA Joint Summary Report. Washington: NASA and Paris: ESA.

Stofan E, Lorenz R, Lunine J, Bierhaus EB, Clark B, Mahaffy PR, Ravine M. 2013. TiME - The Titan Mare Explorer. IEEE Aerospace Conf, Mar 2–9, Big Sky MT.

Young L. 2001. Exploration of Titan using vertical lift ­aerial vehicles. Forum on Innovative Approaches to Outer ­Planetary Exploration 2001-2020, p. 94.


[1]­ radioisotope-generator-asrg/


[2]  There is reasonable evidence that many planetary moons in the outer solar system—including Callisto, Enceladus, Europa, Ganymede, and Titan—are “Ocean Worlds” harboring interior liquid water oceans.

[3] php?articleid=36


About the Author:Elizabeth Turtle is a planetary scientist at Johns Hopkins Applied Physics Laboratory (JHUAPL) and Dragonfly principal investigator. Ralph Lorenz is a planetary scientist at JHUAPL and Dragonfly mission architect.