Historic Aviation Milestone Reached Near Silicon Valley
The Story of Reliable Robotics
By Annamarie Buonocore
With all things considered, it wasn’t that long ago when the Wright Brothers took their first flight in Kitty Hawk, North Carolina. The achievements of aviators over the last century have been wondrous, leading to a great industry full of American patriotism. In 2024, we entered the age of artificial intelligence – a new industrial revolution – that will take the world to the next level of technology. Over the last couple of years, we have heard about chatbots and self-driving cars. Some have dared to dream bigger and think about self-flying planes. A Silicon Valley-based company recently flew a self-flying Cessna 208 Caravan from the Hollister Airport in Hollister, Calif. History was made in November of 2023 when the aircraft was piloted from the ground 50 miles away. Could this revolutionize flying in the long term? We interviewed Robert Rose, the CEO of Reliable Robotics, to learn more.
In Flight USA: How did you end up in the aviation industry in the first place?
Robert Rose: I had two grandfathers who flew in World War II. Aviation was on both sides of the family. One was an instructor pilot, and the other was deployed to fly over Italy and Southern Germany. The latter flew transport planes in Korea, so I heard a lot of flying stories growing up. There was always this expectation that I would be a pilot one day too. My father was an aircraft maintenance officer in Vietnam. My mother was also involved in aviation and got her pilot’s license when I was in high school. Actually, flying with my mom is where I got the most flying experience. She taught me how to fly and gave me an appreciation for flying. We did Civil Air Patrol together, and I was going to take flying lessons, but it’s an expensive thing to do for a kid, so I put that dream on hold. It wasn’t until much, much later in my career that I got back into flying lessons. But I was really struck by the lack of automation in these systems and how much human control they really required behind the scenes. I started doing research during the evenings and on the weekends to find out why we didn’t have more autonomy in aircraft systems. I asked myself why we don’t have autopilots that can do landings, taxis, and takeoffs. It was in researching the answers to those questions that I decided someone needed to start a company that focuses on automation in aviation so that we can improve safety and have more people experience the joys of aviation. Prior to co-founding Reliable Robotics, I worked on the Tesla Autopilot and the Falcon 9 spacecraft, which are also automated systems.
IF USA: Are you the sole founder of Reliable Robotics or are there several founders?
RR: My main co-founder, Jureg Frefel, and I worked together at SpaceX. I have known him for 13 or 14 years. At SpaceX, I managed the software development team, and he managed the avionics computing team. His team built the computers, and my team did the software. There was always a lot of back and forth on why they couldn’t build faster computers and why we couldn’t write better code. We met in the middle, and it all worked out well in the end. That is how we came together to start Reliable Robotics.
IF USA: How long has Reliable Robotics been around?
RR: We officially started in the summer of 2017.
IF USA: Do you like being based in Silicon Valley?
RR: Yes, it has been great. Jureg was down in Southern California, and I convinced him to come up here. Before all of the tech and silicon companies were here, there was a lot of aviation in the South Bay. There is a long history of aircraft development and testing over here at Moffett Field and NASA AMES research facility. There are not a whole lot of places in the world where you have a high concentration of people who know how to build aviation systems. You have pockets in LA, Seattle, and then the Bay Area. Wichita is good for general aviation manufacturing, but if you want to combine it with the technology know-how, those west coast cities are basically your choices.
IF USA: Which aircraft have you automated, just the Cessna 172?
RR: We’ve automated two aircraft actually. We started with the Cessna 172 in September of 2019, and it took several months until we were comfortable flying it without the pilot on board. Then we shifted gears and focused on the Cessna Caravan, the 208 B. The Cessna 208 program began right around the same time we flew the Cessna 172 uncrewed in 2019. We did our first fully automated landing with a pilot on board the Cessna 208 in 2019. After years of safety research, it culminated with this recent November flight of the 208 where we were comfortable taking the pilot out of the plane and remotely operating the aircraft from our control center 50 miles away. In the control center, you can hear all of the things the aircraft can hear on the radio. If another airplane calls out their position on the radio, you hear it. You can see on the screen where that other airplane is. You can communicate with them as if you were in the plane.
IF USA: Do you plan to do this with other aircraft? Electric aircraft?
RR: Yes, the technology is adaptable to many kinds of aircraft, and we plan to license it out. It can certainly be done with electric aircraft. We would just have to customize certain things, but that is definitely in the plans.
IF USA: Tell me a little bit about how it works from the control room. What are the remote pilots doing to make this work?
RR: Sophisticated jet aircraft of today have flight navigation systems, and I realize these systems are not perfect. They allow you to program the flight trajectory. The concept is important because we’re building on that concept. We have built a flight-management system that allows you to tell it where you want the aircraft to fly. It knows taxi, takeoff, and landing. You’re essentially programming waypoints, speeds you want the aircraft to be at during those waypoints, and things of that nature. In the control center, that is the interface we’re giving you as the remote pilot. You’re using a flight-management system. We also give you a headset, so you can record your voice, and it is broadcast outside the aircraft. Surrounding aircraft don’t even notice the difference. You sound like any other aircraft. You’re just not sitting in the plane, and you don’t have the engine noise drowning out your microphone. In the control center, you can hear all of the things the aircraft can hear on the radio. If another airplane call out their position on the radio, you hear it. You can see on the screen where that other airplane is. You can communicate with them as if you were in the plane.
IF USA: Do you contract with any specific companies to build systems for them?
RR: I cannot mention specific relationships, but we do have partnerships with manufacturers, airline operations, and component manufacturers. We can’t do this all ourselves, so we need to partner with aircraft manufacturers, communication system developers, motor developers, and other vendors. We’re pulling that all together, and the main thing we’re adding is the certification basis for how you tie them all together.
IF USA: How do you envision the future workforce for piloting? Will it be like Uber where people can take jobs as they please?
RR: Let me just say that we’re developing the technology for this, and we plan to license it out to companies who will handle the hiring, so I don’t have any plans for that myself. Operators can choose how they would like to operate it. Our system is designed to be operated from a control center. For the foreseeable future, pilots in the operating center will have to be certified by the FAA. I would not call this a gig job. This is not something that you are going to do from home. You will have to go through extensive training, similar to what a commercial pilot has to go through.
IF USA: So they would have to have a pilot’s license.
RR: Yes, they would.
IF USA: How will this affect air traffic control?
RR: It wouldn’t affect their operations at all. They’re listening to the radio, and they hear us talking on the radio. Pilots in our control center have a push-to-talk button just like you have on an airplane. The system is digitizing your voice, and the aircraft decodes the digital message. It gets broadcast out over the VHF radio. You can reach the surrounding miles they would reach. It’s just like sitting inside the cockpit. For ATC, it’s nice because there is no change to the operation. It follows all the same procedures as other aircraft. The one nuance, and I think this is an advantage, the remote pilots can call ATC on the phone. They can talk to them that way. That is a hard thing to do if you lose radio communications inside the aircraft. Because you programmed the aircraft, you can tell ATC exactly what that aircraft is going to do. You can give specific times and altitudes, and I see this as a great advantage. In the regular system, ATC has to make assumptions about what you might do. It is a lot of work on their part to insure this does not lead to airspace conflicts.
IF USA: What was the historic flight experience like for you as the founder?
RR: Yes, we flew the plane from Hollister Airport, just south of San Jose. The pilot, Danah Tommalieh, ran the operation from Mountain View, which is about 50 miles away. She was communicating with the aircraft over satellite. It was pretty uneventful, and that is exactly what we wanted. We had practiced for years putting this together and had run hundreds if not thousands of simulation sessions with Dana to make sure she was prepared to handle the operation. We ran millions if not billions of simulations of the flight control system. We wanted to make sure it could handle all of the conditions it could encounter. We had many emergency response procedures planned. We had redundancy in the system switched to an entire backup channel. We didn’t need to use any of that. Everything went completely normal, and that was what we had hoped for.
IF USA: How many people have been working on this? How many are remote pilots?
RR: We have two people who are remote pilots. In our company overall, a quarter of our employees are pilots. A lot of people over here are excited about aviation. We have about 100 employees, about half of those are based in the Bay Area and centered around Mountain View.
Silicon Valley-based Reliable Robotics recently flew a self-flying Cessna 208 Caravan from the Hollister Airport in Hollister, Calif. History was made in November of 2023 when the aircraft was piloted from the ground 50 miles away. (Photo courtesy Reliable Robotics)
IF USA: What are your goals for next year at this time?
RR: I think it’s still going to take time to get everything certified. We have a long way to go. Our certification plan has been accepted by the FAA, and that took over four years. That was a significant achievement. Now we have to deliver on this system, and that is going to be a lot of work. We’re working to get this system into a number of Caravans across the United States. Then we plan to expand into other aircraft. I think the system has great applicability to many cargo and passenger aircraft. The military is very interested in what we’re doing. We’re studying how to adapt the system to large, multi-engine jet aircraft. This would be a game changer for the Air Force. I think it’s going to improve safety and flexibility of aircraft. Our system will ultimately allow many aircraft to become automated.
IF USA: Do you think you will get into rotorcraft or helicopters?
RR: It could happen in the future, but at the moment, we are focused on fixed wing. It gets kind of technical, but there is nothing about our system that precludes us from getting into rotorcraft in the future. We will consider it, but we’re not actively working on it right now.
IF USA: Do you consider yourself teaching the AI how to fly?
RR: “Teaching” is a strong word because this is not AI. Really, what we’re doing is codifying the rules that people use to fly aircraft. For example, when there is loss of engine thrust, you have to learn the best glide ratio. As a human pilot, you learn this in books and by watching videos. Then you train and demonstrate it to a CFI and DPI that you can respond quickly enough. In an automated system, demonstrating it once is not good enough. It means describing the situation in great detail and then writing code to ensure it can respond correctly each and every time. It takes a lot more work to automate a system than to teach a human. Human beings are remarkable in their adaptability. But on the flip side, we are not perfect, and sometimes we get into situations that can be hazardous.
IF USA: The drone community talks about flying passengers in larger drones. Do you see yourself in competition or integration with them?
RR: People have to understand that there is a great difference between the regulatory environment for small drones and the regulatory environment for aircraft. There is a huge step in the processes and the level of rigor to put an aircraft that can carry people into the sky or any type of aircraft that can carry more than a few thousand pounds. There are big differences in the amount of weight aircraft can carry. I never subscribed to the belief that anything in the small drone community would become applicable to larger aircraft. There might be some airspace issues that might be applicable, but the reliability and safety issues are totally different, and that has to be integrated from day one.
IF USA: Do you think all planes will be pilotless in the future?
RR: I would not say pilotless because we still have pilots, they’re just in control centers instead of in cockpits. There is still great oversight needed due to the way the airspace system works. Due to the nature of evolving emergencies, there will always be situations where you want a human being involved in the decisions that could result in a hazard. If you have to choose between two bad things, you will want a human being involved in that decision. People need to understand how long it will take to make this commonplace. Aviation does not have a reputation for doing things quickly. It will not happen overnight. It just doesn’t work that way. It will take many years.
IF USA: Thank you!
