The recent news that Reliable Robotics has completed a $100 million funding round gives a significant boost to the Mountain View, California company’s efforts to develop automatic systems that it says will make flying safer and eventually remove pilots from the flight deck. Pilots will still be involved in the operation of the aircraft, initially Cessna Caravan single-engine turboprops, but they will be based on the ground and will simply be available to assist when needed.
“Progress has been steady,” said Reliable Robotics co-founder and CEO Robert Rose. Earlier this year, the company demonstrated the remote operation of a Caravan (carrying a safety pilot on board) 50 miles from its Mountain View control center. The company has grown to more than 70 employees from 40 last year.
“The main thing we need to do first is get certified with the FAA,” Rose told FutureFlight. Last year, Reliable Robotics projected certification in two years, “and we’re on track.”
To gain experience, Reliable Robotics obtained a Part 135 charter operator's certificate and began flying passengers earlier this year in a Cirrus SR22. The company plans to start flying cargo in Caravans in the first half of 2022 in the traditional fashion with one pilot at the helm. Jeff Drees, who recently joined Reliable Robotics, brings years of experience in air cargo, having been co-owner and chief commercial officer for regional air cargo carrier Ameriflight. "Automated aircraft present a massive opportunity to transform the cargo industry over the next several years," Rose said.
Once elements of the Reliable Robotics automation system achieve certification, those will be added to the cargo Caravans, he explained. The Part 135 charter operation is based in the Boston area and overseen by the FAA’s Boston Flight Standards District Office, although the decision on where the Caravans will fly has not yet been made, according to Rose.
In the first stages of its automation plan, the Reliable Robotics system will operate the airplane, while a safety pilot flies along to handle any contingencies such as an engine failure. “The first thing we’re doing is an advanced autopilot,” he said. This will be capable of flying an entire mission with the pilot not touching the controls and meeting Part 23 certification reliability standards. If there were a major problem—for example, total engine failure—the pilot would shut off the autopilot, take control, and find a place to land.
In the second phase, the system will be designed to handle contingencies such as engine failure and flight into icing conditions without onboard pilot intervention. “This system would present a resolution to the pilot or identity a resolution and do it,” Rose said. But this might mean not only handling a situation but preventing it in the first place. For example, in the case of icing conditions, the autopilot might replace possible poor pilot decision-making that could have resulted in flying into icing conditions with sensor readings and online weather information that the autopilot can use to find a route free of icing.
“We see automation as a huge opportunity to prevent this situation,” Rose said, “given the amount of weather data remote pilots will have access to and the sensors on the aircraft. It will be continually reviewing the output of those sensors. We should be able to reduce incident numbers.”
Once the automation is fully proven in the first two stages, Reliable Robotics will move the pilot out of the aircraft and to the control center and the aircraft will fly on its own but supplemented by the ground-based pilot. In this case, the remote pilot will program the Caravan’s flight management system (FMS) with the instrument flight rules (IFR) plan as well as backup trajectories in case they are needed. During the flight, if controllers assign a new route, the remote pilot will be able to send it to the airplane’s FMS. The remote pilot will also communicate with controllers through a satellite communications link between the airplane and the Reliable Robotics control center and then via the aircraft’s radios to the controller.
Meanwhile, for the certification effort, Reliable Robotics engineers are working on issue papers that address FAA requirements. “We’re doing something pretty new,” Rose said. This includes developing algorithms, sensors, and avionics in-house that improve integrity monitoring for navigation, such that the Caravan will be able to land in zero-zero weather conditions. The company’s certification efforts are overseen by the FAA’s Los Angeles Aircraft Certification Office.
If what Reliable Robotics is doing pays off, transportation of cargo, and eventually people, will become vastly more efficient, Rose explained. In the near-term, autonomous operations will be similar to the way they are conducted today, with one or two pilots flying each aircraft, making the operation dependent on human duty time limitations and their ability to move around constrained by the network itself. “But there are efficiencies in how we can increase utilization,” he said. “It’s not just about cost savings, it’s about what we can do with the system that we couldn’t do previously.”
What this means is finally putting valuable airborne assets to more efficient use, Rose explained. “They are underutilized now. It goes back to physically the pilot has to be stationed with the aircraft. When you decouple that, it opens up a bunch of possibilities to add more aircraft, perhaps flying at lower frequencies. All these cargo routes are established months in advance. It’s difficult to change those due to the human constraints of having a pilot tied to the aircraft asset. [With autonomous aircraft], you can create and destroy routes or introduce more vehicles.”
Rose cites as an example the “snowmaggedon” disaster in Texas in February, when boxes piled up in distribution centers, unable to be delivered. “Once the runway thawed, they were only able to get packages out at a fixed rate. They were constrained by a route structure put in months prior.” But an automated aircraft—and this is looking further into the future—could analyze the weather situation and figure out which runways would thaw first and dispatch itself to airports just as they are opening and move those boxes. “That’s not feasible the way we’re set up today,” he said.
In the current system, an operator has a fleet of aircraft available to be flown by human pilots. The pilots have physical limitations: they can’t fly 24/7 without rest, they can’t easily switch from one aircraft type to another, and they require frequent training and monitoring of their health.
With autonomous flight, the operator can easily deploy the aircraft when and where they are needed, without having to figure out how to position pilots to be available, well-rested, and safe to fly. Recent disruptions at some airlines, where flight crew are barely able to keep up with heavy workloads, are another example of this problem. Airlines have had to cancel many flights because they can’t get pilots and flight attendants where they need to be at the right time, a true logistical nightmare.
Another advantage, Rose pointed out, is that the operator could field a more diverse fleet, aircraft that meet the needs of the customer, without having to worry about differences training or sending pilots for a new type rating. “It’s hard to have pilots training to fly different types,” he said.
As propulsion system technology transitions to electric motors driven by batteries, hybrid-electric generators, or hydrogen, aircraft operators might choose to replace existing turbine engines with new power systems. If these aircraft are already equipped with autonomous controls, changing the power plant won’t impose a new training burden, except for some training needed for ground-based pilots. “We see autonomy as the stepping stone to get into that faster,” he said.
Ultimately, the pilot-managers on the ground should be able to oversee more than one aircraft. “Longer term, there are also benefits of economies of scale by having a collection of pilots in control centers capable of operating larger sets of vehicles,” he said.
Long term, Rose believes, autonomous aircraft will offer expanded career opportunities in aviation. There will always be a need for pilots to fly non-autonomous aircraft, but as autonomy drives growth in aviation, he said, “[companies] will be operating aircraft at orders of magnitude more than today. In the future, pilot training will look more like air traffic controller training, a collection of people monitoring aircraft. There are a huge number of skills that pilots have today and that will carry over directly [to autonomous operations].” These include preflight weather briefing, flight planning, communication monitoring, aeronautical decision making, and IFR procedures, all of which will carry over directly to remote operations, Rose said. “I think the only thing that is really going to change is [the need for] stick and rudder skills.”
Rose believes that autonomous technology “can definitely improve the safety record of general aviation.” The feeder cargo market is the logical first step for Reliable Robotics’ efforts. “But longer term, we would like to adopt this technology for regular general aviation aircraft so others could use it. That’s the reason I started the company. I love being in airplanes, I love flying,” he said, but like many pilots, he has a hard time staying current enough to feel comfortable taking his family on a trip in a light aircraft. Technology that makes flying easier and safer would enable such trips.
“If you want more pilots to take advantage of the National Airspace System, you have to alter the human-machine interface between the pilot and the aircraft. [That means] building a fully automated aircraft and looking at ways to bring the person back in.”