In September 2019, a Cessna 172—a four-seat, single-engine plane that is among the most common aircraft model in existence—taxied to a runway at an airport south of San Jose, took off on its own, flew for 15 minutes, and then landed at the same airport, all without a single person on board.
The aviation milestone—believed to be the first complete civilian unmanned flight over a populated area—was revealed in a video released today by Reliable Robotics, a Mountain View, Calif., startup founded by two former employees of SpaceX and Tesla that emerged from three years of “stealth” operation.
The flight, for which the Federal Aviation Administration had cleared the company back in December 2018, came several months after the August 2019 flight of a different unmanned small Cessna model, a 1968 Cessna 206, conducted by the U.S. Air Force Research Laboratory and the technology company Dzyne Technologies. But that flight took place over the remote Dugway Proving Ground in Utah, not a populated area.
Reliable Robotics also announced that it has now installed the same autonomous-flying software on a larger single-engine Cessna 208 Caravan—which is often used for short-haul cargo deliveries and passenger flights—and, in June, landed that airplane autonomously for the first time.
The revelations come a week after another California-based autonomous aviation company startup, XWing, also came out of stealth mode and revealed that it had conducted the first full autonomous flight of a Cessna 208 Caravan, although in XWing’s case, a human pilot was onboard ready to take back control of the plane in the case of an emergency.
Reliable Robotics is the brainchild of Robert Rose, who was once director of flight software at Elon Musk’s rocket company, SpaceX, and later developed autopilot technology at Musk’s Tesla, and Juerg Frefel, who designed the computing platforms used for SpaceX’s Falcon 9 launch vehicle and Dragon space capsule.
The two founded the company in 2017 with the vision of trying to get airplanes flying autonomously as quickly as possible.
One of the biggest impediments to self-flying planes is regulation, so the two founders tried to figure out how they could convince the FAA to approve unmanned flights without having to modify any existing regulations.
Their conclusion was that while most other autonomous-flying companies are pursuing brand-new aircraft designs that often look like oversize consumer drones—with four or more electric rotors that allow the aircraft to take off and land vertically and hover in place. It would be easier to gain FAA certification if they simply converted a tried-and-true existing airplane for robotic flight.
The Cessna 172 debuted in 1955 and the Cessna 208 was FAA certified in 1984. They have well-known safety and reliability records. This way the government agency would have to certify only the company’s software and safety plans, not the airframes and engines.
Rose, who is Reliable’s chief executive officer, says the company’s business plan is to equip a fleet of Cessna 208 Caravans with its self-flying systems and use them for air freight deliveries. He predicts the company will be able to begin commercial autonomous cargo deliveries in the U.S. within two years.
Rose says the company’s ultimate goal is to fly passengers autonomously too, but getting FAA approval to do so will take more time—and probably will require additional rule-making by the agency.
This plan closely tracks that of Reliable Robotic’s rival XWing, which also plans a fleet of autonomous Cessna 208 Caravans for hauling cargo as a stepping-stone to passenger flights. The company, which has purchased a small Texas air freight company to acquire an FAA air cargo license, says it plans to begin real cargo delivery flights “within months” to further hone its systems, although these flights will also carry human pilots for safety.
XWing has raised about $14 million in venture capital to date. Reliable Robotics has raised more than twice that amount—$33.5 million, including a $25 million Series B round in March 2019 lead by Eclipse Ventures, which was also publicly announced Wednesday. Reliable Robotics employs about 35 people.
So far, Reliable has had to have an observer on the ground directly watching its airplane fly because of FAA rules that stipulate civilian drones must be flown within direct line of vision of a human operator. But Rose believes this regulatory barrier can also be overcome to let autonomous aircraft fly far greater distances. “I believe it can all be done within the existing regulations,” he says.
Rose says that when the company begins autonomous cargo flights, its aircraft will have a human pilot monitoring them from a ground control station and talking to air traffic control. This pilot will use Reliable’s software to issue high-level commands to the aircraft, like hitting a button that tells it to complete a takeoff sequence, for instance, rather than directly manipulating the airplane’s controls from afar, as is the case with most drones today.
He says this will enable a single pilot to conduct many more cargo flights per day than is currently possible, significantly reducing costs for air freight carriers. “Today, with the concept of operations and the patterns in which they fly the aircraft, the pilot is grossly underutilized,” he says.
Billionaire venture capitalist Peter Thiel famously knocked the level of ambition in Silicon Valley with the line, “They promised us flying cars. Instead, we got 140 characters.”
But in the past five years, there’s actually been a lot of companies trying to build flying cars. They include startups like Kitty Hawk, which is backed by Google founder Larry Page, as well as Skyrise, and two German startups, Lilium and Volocopter, that have attracted plenty of venture funding, not to mention Uber, as well as aviation industry giants like Airbus, Boeing, and Sikorsky.
But the majority of these companies are working on brand-new, often exotic, aircraft designs. And while most of these companies say they eventually hope to create vehicles that can fly themselves, without a human pilot, they are starting off with a human in the cockpit to help minimize the risks involved with developing experimental new aircraft and getting government approval to carry passengers.
Another group of companies are working on autonomous drones to solve “last mile” delivery of packages. These include startups like Zipline, which has used drones to deliver medical supplies to remote spots in Africa, as well as behemoths like Amazon, which has built delivery drones that can carry payloads of up to five pounds as far as 15 miles, as well as rival delivery companies, like UPS and FedEx. There’s even another subset of businesses, such as San Francisco startup Elroy Air, that are designing new, massive vertical takeoff and landing drones that can lift far heavier cargos and carry them much further.
In many ways, what Reliable Robotics and XWing propose to do with unmanned flight is less revolutionary. But as a result, it might just have a better chance of becoming reality in the near term.
