How a Smartphone Chip Powered NASA’s Historic Flight on Mars

On Monday, an aircraft flew on another planet: a historical first. Here's a look at the tech that made it happen.
article cover

Francis Scialabba

· 10 min read

For six decades, humans courted the Red Planet—and as tech advanced, so did the Martian milestones. The first flyby mission? 1965. First landing? 1971. First rover? 1997. But the public hasn't had a big-time mission to celebrate in more than twenty years.

That is, until this week.

On Monday, NASA’s Ingenuity helicopter made history as the first-ever aircraft to take flight on another planet. Call it one small step for a 19-inch robotic helicopter, one giant step for humankind. It was accompanied by the Perseverance rover, which helped act as a communication guide for the $85 million helicopter’s autonomous remote journey—a 40-second flight through Mars’s thin atmosphere, signaled by a crew 181 million miles away. The mission paves the way to explore previously inaccessible features of Mars’s terrain: canyons, volcanoes, craters, and more.

“Many of us have put the last five years of our lives into this, and so actually seeing it happen—and knowing it was the first time it ever happened like this anywhere in history—it was very exciting to be a part of that,” Timothy Canham, senior software engineer at NASA’s Jet Propulsion Laboratory (JPL), told Emerging Tech Brew.

The most important interplanetary development in two decades hinged on a host of familiar emerging technologies, from chips and battery tech to robotics, drones, and autonomous flight algorithms. One of the most mundane, but critical, building blocks is the Qualcomm chip inside the helicopter and aboard the rover—the very same semiconductor tech used in 40% of the world's smartphones. Ingenuity’s flight is a tale of a technological breakthrough made possible by decades of unassuming, and seemingly unrelated, building blocks.


From smartphones to space machines

If you’re an Android user, Snapdragon 801 is probably the chip responsible for making your smartphone work. So how did the semiconductor powering green texts worldwide make its way to Mars?

Qualcomm’s internet of things division started working on adapting smartphone chips for drone use on Earth in 2012, as part of its Qualcomm Flight platform. The goal: Use Snapdragon to power smaller, lighter drones, with onboard sensors, AI, navigation, and cameras.

These terrestrial drones are used in acid detection, delivery, search and rescue, and more. Their jobs require vastly different skills than a smartphone’s: navigate without a GPS, do a lot of computing, and fly autonomously. So Qualcomm built an entirely new flight platform around the Snapdragon processor, including a real-time robotics operating system, a camera-based vision system, and algorithms for functions like obstacle avoidance, flight planning, and visual inertial odometry, which helps a machine “position itself accurately even in GPS-denied locations,” Dev Singh, GM of robotics, drones, and intelligent machines at Qualcomm Technologies, told us.

In 2014, Qualcomm’s CES presentation on Snapdragon-powered drones caught the attention of NASA’s JPL, which was trying to prove a helicopter’s flight viability on Mars. The lab bought a Qualcomm Flight platform and Snapdragon chip from Qualcomm and started experimenting.

“Most space electronics are big, heavy, and bulky,” says Canham, and JPL was looking for something very different here: a small, lightweight, and compact—but computationally powerful—platform that could help them reach a historical milestone. After a 2016 study, the lab landed on Qualcomm’s as the best option due to its combination of size and power. Plus, it came with compatible cameras, which could save JPL’s small team several months of development time.

In 2017, JPL heard that the Qualcomm Government Technologies team happened to be in the NASA building and asked them to stop by. That turned into a day-long meeting: the team behind the smartphone chip and the team readying it for use in space, sharing obstacles and brainstorming solutions.

There were ultimately no hardware changes to the Snapdragon chip, but NASA did give it a partner: a second chip, an automotive-grade microcontroller, that was rugged enough to withstand a vehicle’s high temperatures. Snapdragon took care of image processing, guidance processing, and storing flight data—with readings 500 times a second—while the microcontroller was in charge of navigation and running the helicopter’s motors.


Using IO connections—a form of communication often used by tiny devices—the two are able to communicate with each other and the Perseverance rover’s “base station” via radio, which also runs on the exact same Snapdragon chip. Since Perseverance can’t talk directly to Ingenuity (it’s a different kind of radio, says Canham), the rover’s base station serves as the middleman. Perseverance sends command files from Earth and the rover to Ingenuity, and Ingenuity sends flight data back to the rover, which then downlinks to Earth.

“A one-way radio frequency message coming from Earth to Mars is anywhere from 13 to 15 minutes” to arrive, says Chris Pruetting, a senior director of business development at Qualcomm Government Technologies. “We talk often here on Earth about processing at the edge, what our cell phones can do today. We talk about cloud-based computing. Some of these things are just kind of taken away from the arsenal when we talk about, ‘How are you going to fly on another planet?’”

While the hardware stayed largely the same, NASA had to build software from the ground up. Qualcomm’s original flight platform was intended for this planet, not a trip to Mars.

For years, a team within JPL worked on specialized flight algorithms that would tell Ingenuity how to operate in an incredibly thin atmosphere, just 1% of the density of Earth’s, and a Mars flight simulation environment. The robotic helicopter has autonomy, meaning a team on Earth sends it a flight plan (high-level commands like where to fly and when to land), and Ingenuity’s algorithms figure out how to make that happen.

Keep up with the innovative tech transforming business

Tech Brew keeps business leaders up-to-date on the latest innovations, automation advances, policy shifts, and more, so they can make informed decisions about tech.

The helicopter’s software needed to “work hard and fast to keep itself aloft,” says Canham, constantly measuring the sensor inputs and making hundreds of decisions each second to guide the helicopter’s 1.2-meter-wide blades. That includes spinning them at 2,500 rotations per minute to stay airborne (comparatively, a typical Earth helicopter would be 500 or 600 rpm).

JPL also had to figure out how to operate all of this amid radiation, wind, shock during takeoff and landing, and extreme cold—down to -99 degrees Celsius at night. One fix: Installing a heater on board to keep the electronics warm. But they had to balance battery use at night with the battery life needed for daytime flights.

Then there’s the issue of helicopter navigation on Mars. It requires “a lot of triangulation of data,” says Singh. “It’s not an easy problem.” For a machine like Ingenuity, sensor inputs, cameras, inertial IMUs, a gyrometer, an altimeter, temperature sensors, and more “collectively feed into the algorithm” to give the machine “sight,” says Singh.

Before sending Ingenuity to Mars, JPL ran extensive tests—pumping down the air pressure in special chambers to approximate Mars’s surface, using a “gravity offloader” to tug down on the helicopter similarly to the planet’s gravitational pull.

“We were very nervous,” says Canham. “We didn’t know what was going to happen until we actually got there, because it had never been tried before.”

And then, just before launch, they discovered a bug.

Embedded chips for high-stakes vehicles—planes, for instance—require safety mechanisms or, as Canham put it, a “dead man’s switch.” Think about spy movies, where the deep-undercover agent periodically sends a secret code back to their organization to confirm they’re alive; if, one day, they don’t send the code at the appointed time, then the organization could step in with drastic measures.

For Ingenuity, there’s an equivalent: the Snapdragon processor sending a special pattern to the hardware to confirm it’s working. If something happened to the chip, and it didn’t send the pattern, then the hardware would consider those drastic measures.

Of course, the pattern the hardware wanted didn’t come through—so it wouldn’t allow the processor to operate.

The world was watching as NASA made the announcement: Ingenuity’s flight, initially scheduled for April 8th, would be delayed until the 11th. And then, days later, another announced delay: the 14th. It would finally take off on the 19th.

Although things were “very quiet for a while, in terms of what you were hearing from the JPL press office,” says Canham, he and a hardware designer spent 10 or 15 hours a day, for four or five days, trying to dream up a workaround for the bug they’d discovered in the hardware. Hardware that was already on the surface of Mars.

Luckily, the team found a software workaround for the hardware problem: an “existing backdoor” in the software that allowed them to “‘tickle’ the hardware and force it to reset this check,” says Canham. “It was quite the Apollo 13 moment.”

Just after 3:30am ET on Monday—12:30pm Mars time—Ingenuity took flight.


On the horizon

Ingenuity may be small, but the rotorcraft has enormous implications for our understanding of Mars, our universe, and even our own planet, says Dr. Matthew Shindell, a curator of planetary science and exploration at the Smithsonian’s National Air and Space Museum.

To get an idea of the modern-day limits of a rover exploring Mars’s surface, imagine a heavily fortified Roomba. You’ve got to make sure it’s set down on a flat surface, with no treacherous falls, walls, or major obstacles around it that could get it stuck and end its journey early. Because Mars has a lot of treacherous terrain, “We haven’t actually gotten to study, up close, things like canyon walls or volcanic slopes,” says Shindell.

With the power of flight, that can change. Not only will scientists be able to get up close and personal with Mars’s surface features, but robotic helicopters like Ingenuity will also be able to send entirely new long-range views back to Earth.

“When I think about Ingenuity and think about what it means for the future, I think back to that 1997 Pathfinder-Sojourner mission,” says Shindell. “If we think...the moral of that story [is] that Sojourner led to bigger, better rovers, then I think that is what we might expect to see from the Ingenuity experiment.”

Canham calls that a “very good analogy.” Sojourner’s job was to prove the new technology and report back, and Ingenuity’s purpose is similar: prove the tech and gather data for the next, even more ambitious, try.

JPL is already working on the next generation of Mars helicopter—an upgrade from four-pound Ingenuity to a 50- to 70-pound vehicle, one that can carry an instrument, with even more sophisticated navigation and higher-level autonomy. The team is also developing new algorithms with 3D obstacle detection, which would help steer a helicopter around a boulder, for example, and land it safely. (Ingenuity doesn’t have that capability.)

NASA is “already looking at a lot of these next-generation concepts, where they can take what they’ve learned from Ingenuity and feed it forward to even better helicopters,” says Canham.

Those next-generation helicopters will likely still use the same chips that sit in the pockets of more than a billion people worldwide.

Keep up with the innovative tech transforming business

Tech Brew keeps business leaders up-to-date on the latest innovations, automation advances, policy shifts, and more, so they can make informed decisions about tech.