A California company called Reliable Robotics has achieved some pioneering achievements in unmanned aviation, having already flown a Cessna 172 Skyhawk and a 208 Caravan without pilots in a relatively populated area and normal metropolitan airspace. The flights took place from San Martin Airport, about 25 miles southeast of San Jose, first with an unmanned Cessna 172 on Sept. 24, 2019, followed by the Caravan’s first remote automated landing on June 30, 2020.
While these weren't the first unmanned flights of a passenger-type general aviation aircraft, that Reliable Robotics was able to gain FAA approval to conduct the flights so close to major cities in the San Francisco Bay Area is significant. But there is more to what Reliable Robotics is doing, and this could have a profound impact on the advanced air mobility (AAM) or urban air mobility (UAM) market.
CEO Robert Rose and v-p of engineering Juerg Frefel co-founded Reliable Robotics in 2017. According to the company, Rose “led flight software at SpaceX and the autopilot program at Tesla,” and Frefel “led the team developing the compute platform for the Falcon 9 rocket and the Dragon spacecraft.” Since it’s founding, Reliable Robotics has raised $33.5 million in two rounds of funding ($8.5 million Series A led by Lightspeed Ventures and $25 million Series B led by Eclipse Ventures with participation by Pathbreaker Ventures and Teamworthy Ventures).
It may seem unlikely, but what Reliable Robotics is striving to accomplish isn’t to design and manufacture the next great unmanned aerial passenger-carrying vehicle but something more critical, the technological underpinnings it says will make it possible for such vehicles to succeed. The products that Reliable Robotics makes will enable vehicle manufacturers to automate the flying, and this could apply to converting an existing aircraft to unmanned operations or adding a Reliable Robotics-designed flight control system to a new design.
The long-term goal, Rose told AIN, is to make more efficient use of the 5,000 public-use airports in the U.S., most of which have a lot of underutilized capacity and are located near where people live. Most people, he explained, might see about 30 of the largest and busiest airports during their travels, while fewer people might occasionally use another 170 smaller regional airports. However, increasingly, people are moving to desirable locations away from metropolitan areas, and the thousands of available airports could be a tremendous resource. “We believe automated aircraft are the key to unlocking these airports for use by regular people,” he said.
Rose is aware that people are still wary of automated vehicles. “It will be quite some time before people are comfortable and the technology is ready to be flying passengers around, but before that, there is a lot of opportunity in the cargo space,” he reflected.
Certainly, many charter companies have tried to crack the nut of serving far-flung passengers. "If you’ve got a lot of money, you can do it,” Rose said. But the need for pilots to be available when passengers want to fly hinders the growth of these companies. “When we’ve broken this market down," he said, "a lot of the reason you don’t see this is pilots don’t live next to the airplane or where the demand is.” A huge amount of charter flying is just to reposition the aircraft for the next passenger flight and this adds more cost and logistical challenges for matching available pilots to the aircraft.
“When you automate the aircraft operating model for these kinds of businesses, it completely changes,” he said. Unmanned aircraft would cost less to operate, and what Rose envisions as remotely piloted operations have no need to have a pilot available to fly inside a particular aircraft. Even if passengers insisted on a live pilot onboard, this could be accommodated in the early stages, with higher efficiency obtained by flying the deadhead (passengerless) trips unmanned. The unmanned aircraft would also be ideal for cargo operations.
Rose believes that unmanned aircraft will have a big impact on safety. “In my limited aviation experience, I’ve been astounded at how complex flying a plane is, especially when things go wrong,” he said. “The basics of keeping it in the air requires a great deal of cognitive capability. An unmanned aircraft takes care of flying,” he explained, “and frees a person up to think how to get out of a situation or solve a problem.” If that means the aircraft is guided by a remote crew on the ground, all the better, because at the control center, he said, “We have a lot of people there ready to assist in in the event something does go wrong. We’ve got a room full of experts, a brain trust of people who know how to operate this plane.”
The system that Reliable Robotics has designed for the Skyhawk and Caravan operates in parallel with the existing flight controls and it can be switched on and off in case a pilot needs to fly the airplane. For example, after heavy maintenance, a pilot could still conduct a test flight, or someone might need to move the airplane on the ground by taxiing it.
The new system “is tying into the [control] cabling like an existing autopilot would,” Rose said. “We use much stronger, more capable actuators that have greater redundancy and fault management and fault isolation capability.”
The aircraft's existing autopilot wasn’t usable, however, because it isn’t capable of autolandings and the navigation sensors in the existing avionics aren’t precise enough, Rose explained. “We’ve developed a navigation stack that allows us to precisely position the aircraft in the sky or on the taxiway.”
Although he wouldn’t provide too much detail, Rose said, “We have borrowed some techniques learned from the space industry and applied them to aviation. They’re not particularly novel. We’re able to utilize these techniques to show the integrity of navigation. It’s a combination of sensors. That’s where it is novel.”
The concept of a flight management system (FMS) isn’t new to aviation, but Rose still views the various inputs as separate (VOR, DME, GPS, etc.) “The pilot acts as the integrator,” he said, which is true in that the pilot has to interact with the FMS to navigate. He added, “We’ve developed another type of system that fuses this type of information to give a single cohesive view [to the aircraft operator].”
With the automation switched on, the aircraft is then connected to the control center through “redundant control links,” Rose said. The links are VHF or satellite and enable real-time control of the aircraft from the ground.
From a regulatory standpoint, Reliable Robotics sees a pathway under the new FAA Part 23 rules, which are intended to accommodate new technologies. The aircraft have to fit within the existing airspace infrastructure, so they are equipped with ADS-B Out transponders and use VHF for communications with air traffic controllers or to broadcast on local frequencies at non-towered San Martin Airport. The communications are done by the pilot-operator on the ground, via the link to the airplane. The aircraft are approved to fly on an experimental demonstration airworthiness certificate.
Rose was inspired to start Reliable Robotics after reading Uber’s seminal Elevate white paper, which laid out grand plans for electric-powered UAM vehicles carrying people through the air instead of on the ground. “I was really inspired by that,” he said. His family had long been involved in aviation and Rose joined the Civil Air Patrol while in college. “When the paper came out, I thought, ‘this is great, I’m going to jump in.’”
At first, Rose looked into designing new aircraft, as did every other UAM aspiring pioneer. Have worked at SpaceX and Tesla, he recalled, “I’ve done vehicle design, and it is a slog. I started looking at what it took to do an [already] FAA-certified aircraft, and this is the slog to end all slogs. But I was enamored by the use case and the potential of automated aircraft.”
When he spoke with the companies in the eVTOL/electric space, he found that their founders were wary of the FAA certification process and thought they might have an easier time seeking certification in other countries, then bringing their products to the U.S. for FAA acceptance. EASA, after all, was the first regulator to develop rules covering eVTOL and electric aircraft. But Rose soon learned that this pathway wasn’t going to make manufacturers’ jobs easier and that working with the FAA early would be far more productive.
So the problem that Rose decided to solve was not how to design a new eVTOL or electric airplane but something that all of the designers and manufacturers of those products would need, the ability to fly without a pilot and the regulatory framework that would allow this. “I got pretty excited about this,” he said. And after all, the SpaceX rockets that he had helped make possible were the ultimate unmanned vehicle. “At SpaceX there was no process to do a commercial launch, and we created it,” he said. “We were the ones explaining to the FAA the process by which they’d certify it.”
Ultimately, he explained, “Somebody needs to do this, and why not us?” The result was Reliable Robotics and the recent culmination of its efforts, the unmanned flights of the Skyhawk and the Caravan.
The next steps are to continue testing, working with designated engineering representatives, and gaining FAA confidence and a clear pathway to certification so customers can start operating unmanned aircraft, at least for cargo transport. In fact, the registration number of the Caravan--N927FE--pretty much reveals one company that is seriously interested in this technology (FedEx), although Rose can’t speak about it due to a nondisclosure agreement. “There is definitely interest by a number of operators in the U.S.,” he said.
Rose is confident that passengers will eventually elect to fly on unmanned aircraft. “I’m optimistic that when automated aircraft become more commonplace, they are going to be in a situation where there will be a mechanical failure or weather event that leaves you stranded, and the choice is to stay overnight or jump in an automated plane that’s leaving that night.”