How to Conquer Titan With a Nuclear Quad Octocopter

The Johns Hopkins University Applied Physics Lab's Dragonfly is a finalist for a NASA mission to Titan in the 2020s

Dragonfly Concept
Image: JHU APL
A concept drawing shows the Dragonfly quad octocopter designed by Johns Hopkins University's Applied Physics Lab.
Advertisement

In December, NASA announced two finalist concepts for a robotic mission that will launch in the mid-2020s. The first is the Comet Astrobiology Exploration Sample Return (CAESAR), which would send a fairly conventional spacecraft over to a comet to grab a chunk of its nucleus and bring it back to Earth. That's cool and all, but we're much more excited about the second finalist concept: Dragonfly, from the Johns Hopkins University Applied Physics Lab (APL), a quad octocopter that would explore Saturn's moon Titan from the air. The idea is that it would work like a planetary rover, except that it would fly instead of drive, allowing it to cover much more ground at the risk of, you know, crashing.

We've seen lots of drones that can do amazing things, and also lots of drones that crash very, very badly while trying to do amazing things. Sending a fully autonomous flying robot to an alien world over a billion kilometers away and expecting it to fly around for a couple years without any human intervention seems extraordinarily ambitious, so we checked in with APL to see exactly what they're working on.

Titan is an appealing place to visit because in some important ways, it's more like Earth than any other place in the solar system. It's a bit chilly, with an average surface temperature of just 94 kelvin, and the atmosphere is thick (four times thicker than Earth's) and made up of nitrogen and methane. It also has one-seventh the gravitational pull of Earth. But what's so interesting about Titan is that it's got a methane cycle much like Earth's water cycle, with liquid methane forming lakes and rivers and clouds and rain. There's also a bunch of organic compounds thrown into the mix, which makes it an intriguing place to look for very primitive, and very weird, life.

The Cassini mission to Saturn included a little probe called Huygens, which was dropped on Titan in January of 2005. Huygens was mostly designed to

measure atmospheric conditions, but it managed to survive landing on the surface of Titan for a little over an hour anyway, and sent back a picture of the surface.

When considering how the surface of Titan might be explored, the moon's exotic characteristics open up many more creative options than would be available for a planet like Mars. In particular, the low gravity and high atmospheric density of Titan favor flight, which would allow an exploration robot to visit many more sites of scientific interest much faster than a stationary lander or a rover. In the past, NASA has considered things like helicopters and hot air balloons and airplanes, but over the last decade, multirotor drones have become the standard for maneuverable and dependable robotic flight. According to APL, Titan is in fact “the easiest place in the solar system to fly a quadcopter.”

Dragonfly is based around a 300-kilogram-ish “quad octocopter” design, which is a quadcopter that has motors and propellers that are doubled up. There's a small aerodynamic penalty for arranging things this way; strictly speaking, a conventional octocopter with eight separate propellers would be more efficient. But Dragonfly has to fit inside a hypersonic aeroshell for delivery to Titan, and the quad octocopter is much more space efficient while still maintaining a generous amount of redundancy.

The reason that Dragonfly is as bulky as it is in the first place is that it needs to bring along its own power source. Titan is far too hazy and far too, uh, far from the sun for solar power to be a viable option, so Dragonfly will rely on the same kind of power system that the Curiosity rover is using on Mars: a radioisotope thermoelectric generator (RTG) that converts heat generated by plutonium-238 into electricity. RTGs can operate for decades and are especially useful for deep space missions because the “waste” heat they generate can be used to keep things nice and warm.

By itself, the RTG can't produce enough sustained power to keep Dragonfly in the air (or whatever you want to call the stuff on Titan that it would be flying through). Instead, the RTG continuously charges Dragonfly's batteries, and eventually, the robot will have saved up enough energy for a flight. Or, maybe it uses that energy to transmit data back to Earth, or do science experiments. Titan's days and nights are eight Earth days long each, so while it's dark out, Dragonfly will have plenty of time to recharge.

Unlike the Mars rovers, which spend a lot of time on the move, Dragonfly would spend most of its time stationary, doing science, transmitting data, and charging. Think of it more like a “relocatable lander” than an aircraft. The science payload would include a mass spectrometer, gamma-ray and neutron spectrometer, geophysics and meteorology package, and of course a comprehensive suite of cameras. Drills on the struts of Dragonfly's landing gear would collect surface samples, and a foldable high-gain antenna means that no relay orbiter is necessary.

While doing science on the ground is valuable and safe, the whole point of Dragonfly is that it can fly. In a hour, Dragonfly will be able to travel farther than any of the Mars rovers could drive in a decade, all without worrying about obstacles. Over a mission duration of a year or two, Dragonfly will be able to explore, in detail, an enormous amount of Titan's surface, and that exploration starts immediately. Rather than being landed from orbit, Dragonfly will power up, drop out of its parachuting aeroshell, and fly itself down, with the ability to scan kilometers of terrain before choosing the safest place to make its first landing. The initial landing area will likely be in large dune fields, which (based on radar imagery) tend to be smooth, with nice flat areas in between the dunes. From that point, Dragonfly can make brief aerial trips to scout terrain from above, but the really exciting part of the mission comes next, as it begins to explore farther and land in new places. Here's how APL says that will be done safely:

  1. A second landing zone (B) is identified by ground analysis of reconnaissance imaging, a distance [range]/3 or less away from the initial landing site A.
  2. The vehicle makes a sortie over this zone using its sensors (lidar for terrain roughness, imaging, etc.) and returns to the original landing site (A).
  3. Analysis on the ground of the sensor data confirms one or more safe sites within zone B (or if no satisfactory site is found, return to step 1).
  4. A candidate third landing zone (C) is identified in reconnaissance imaging, a distance 2[range]/3 away from A.
  5. The vehicle makes a sensing sortie over (C) but lands at (B).

In this way, the mission need not commit to landing sites that have not first been assessed for safety. This conservative approach, while taking longer to achieve a given multihop traverse range, would allow Dragonfly to contemplate much rougher terrains that may be associated with more appealing scientific targets (e.g., cryovolcanic features or impact melt sheets where liquid water may have interacted with organics on Titan).

Dragonfly's maximum range is likely about 60 kilometers, with its battery allowing for a two-ish-hour flight at an optimum speed of about 10 meters per second. In addition to long distance flying, Dragonfly will be able to move itself short distances by "shuffling" across the surface a few meters at a time to reposition its sensors. In terms of the sheer amount of science that Dragonfly will be able to do, its mobility makes it strongly competitive with a more traditional rover, although it makes particular sense for an environment like Titan.

For more details, we spoke with Ralph Lorenz at APL, who helped organize the Dragonfly concept and is the project scientist. 

IEEE Spectrum: What are the flying conditions on Titan like (relative to Earth) and how are they reflected in Dragonfly's design? In what ways would a rotorcraft behave differently on Titan than on Earth?

Ralph Lorenz: Titan's atmosphere is four times denser than ours, and the gravity is seven times smaller, so it's very easy to fly. Actually Dragonfly's aerodynamic design isn't that different from the “quadcopter taxis” being trialed for Earth, but the power to the motors is much less for Dragonfly because of these factors. The denser atmosphere (higher Reynolds number) means we choose a slightly different wing section for the rotors (rather like the ones used on wind turbines).

[In flight] it would basically look the same [as on Earth], except things would look a little bit in “slow motion” on Titan, where the gravity is lower.

How would Dragonfly take off, navigate, and land autonomously? What kinds of sensors would it use?

We can't get into too many details on this, but lidars and radars have been used on terrestrial drones and planetary landers. To a first order, scientists on the ground will tell Dragonfly broadly where we want to go, and it just has to find a flat spot in that general area. And we can always fly back to our first landing spot that we know is safe.

Can you describe the first hardware experiments you're planning on doing? How will you be able to test the hardware on Earth? 

We use a full-up test model (that is less massive than the real thing) and validate numerical models of the dynamics with that.

What kinds of recent technological advances have given you confidence that a robot like Dragonfly could reliably work?

The drone revolution.

Since Dragonfly is still under active and competitive development, Lorenz understandably wasn't able to answer all of our questions with as much detail as we'd like. For example, we're very curious why Dragonfly is using an open rotor design. It's a high-risk, high-reward mission, but there are some ways that such a risk could be mitigated: If Dragonfly had some protective features in common with the kinds of drones that EPFL and Flyability have pioneered, it seems like it might be more likely to survive a bad landing.

The Dragonfly team presumably has a lot of confidence in their approach, which is good, but at the same time, we're talking about a mission that costs somewhere in the neighborhood of a billion dollars and will take decades to carry out. What's the appropriate amount of paranoia for something like that? Although, I guess you could also argue that few things that NASA has done seem crazier than lowering a Volkswagen-size Mars rover from a rocket-powered sky crane, and they managed to make THAT work.

In the spring of 2019, NASA (if we're lucky) will select Dragonfly to continue on to subsequent mission phases. If everything goes well, Dragonfly will launch for Titan before the end of 2025, with an arrival sometime in the mid-2030s.

[ Dragonfly ]

Robotics News

Biweekly newsletter on advances and news in robotics, automation, control systems, interviews with leading roboticists, and more.

About the Automaton blog

IEEE Spectrum’s award-winning robotics blog, featuring news, articles, and videos on robots, humanoids, automation, artificial intelligence, and more.