Earlier this year at the Vertical Flight Society’s 2019 Electric VTOL Symposium, Joby Aviation chief test pilot Justin Paines took the stage to discuss the concept of unified flight control in the context of eVTOL aircraft. A former Royal Air Force (RAF) test pilot, Paines spent years flying and evaluating Harrier jets before joining the Joint Strike Fighter program in 1999, flying all versions of the X-35. After retiring from the RAF in 2004, he continued with the program as a civilian, conducting flight tests that led to the adoption of the unified flight control strategy for the short take-off and vertical landing (STOVL) variant of the Lockheed Martin F-35.
As Paines explained, VTOL aircraft face fundamental flight control challenges because they’re essentially two vehicles in one — a vehicle that hovers and one that flies forward through the air, with the piloting task changing fundamentally between the two regimes. Conventional VTOL aircraft such as Harriers and helicopters are relatively hard to fly. Even after undergoing extensive training, pilots of these vehicles are subject to high workloads that can, and do, lead to cognitive failures and accidents.
The unified flight control strategy aims to simplify the operation of VTOL aircraft by keeping control actions consistent between the two regimes. In the F-35, a left-hand inceptor always provides fore/aft acceleration control. The right-hand inceptor provides vertical axis control when moved fore and aft, and bank angle and roll rate control when moved side to side.
Decoupling the axes means sacrificing the ability to make smooth diagonal control inputs in a hover, which is something that I as a helicopter pilot have always taken for granted. However, Paines said that in practice pilots adjust quickly to making two handed inputs for hover positioning, and the tradeoff results in a vehicle that is dramatically easier to control — which is why Joby is integrating it into the S4, its all-electric (and still largely secretive) CTOL/VTOL aircraft.
After his presentation, I caught up with Paines to learn more about his role at Joby and what’s involved in flight testing a radically new kind of aircraft. (This interview has been edited and condensed.)
eVTOL.com: So from a big-picture perspective, how do you go about testing this brand-new type of vehicle? What’s the approach to that?
Justin Paines: The underlying principle in all flight test is a build-up approach, where you start from a known position, and you build up in whatever parameters you’re expanding the envelope. So, typically a conventional airplane will take off and leave the gear and flaps down, and [the test pilots will] check out the configuration and the response before they raise the flaps and the gear. And then they’ll check out the airplane with the gear and the flaps raised. And when I say check out, you’re looking for a couple of different things to confirm, first, that the airplane’s responding adequately, and secondly, that it’s matching whatever model predictions you have. In the old days they didn’t have model predictions; now they have model predictions.
That’s the underlying principle of all flight tests, and it’s no different in VTOL. You would start from a stationary low hover and check out the performance of the vehicle in the hover and make sure it matches model predictions and make sure your stability margins are sufficient, that you have sufficient handling qualities, response, no delays, all those sorts of parameters. And then you’d increase your airspeed up to 10, 15 or 20 knots, whatever your aerodynamicist said was a reasonably small step in the forces of the airplane changing, and you’d check it out at that speed. And if everything matched there, you’d go to the next speed.
eVTOL.com: Are you able to speak at all about how you’ve approached it with [the Joby S4] specifically?
JP: I don’t think I’d better, because I’d probably get in trouble! But I can say that we are starting in the hover. We regard that as clearly the safest approach for our kind of vehicle. To start at zero knots and “zero” altitude, or five feet, and expand that out from there.
eVTOL.com: So with respect to figuring out the flight controls for this vehicle, is that something that you’ve worked with the engineers with from the beginning on, to provide your input as a pilot?
JP: I’ve only been with Joby since September, and the company was already looking toward unified control, and they probably wouldn’t have hired me if they weren’t. But yes, the unified flight control system is a joint effort between the test pilots and the flight controls engineers. Because they don’t just build the flight control system and then you go and fly the airplane; simulation is a big part of flight control development. So they’ll develop the flight controls a bit at a time, sometimes a piece of the envelope at a time, and we’ll do simulation evaluations and we’ll feed back what’s good and what’s bad and they’ll iterate the flight controls on the basis of pilot feedback. So test pilots, they spend a lot more time in the simulator than the aircraft.
eVTOL.com: So as you’re developing these flight control principles, how much can you do in the simulator before you just have to get in the aircraft and see and adjust?
JP: You do a lot in the simulator, but the final proof of the pudding is when you get to the aircraft. There’s always going to be some at least minor differences —well, there are big differences. In the simulator you will deliberately tune the flight controls to feel pretty heavy and sluggish, because what you find is if you tune the simulator to get the right kind of response that you like, when you get to flight it’ll feel very abrupt, very over-sensitive. So there are differences, but we know about those differences, and we deliberately tune for a slightly sluggish feel in the simulator to hopefully get it right in flight.
eVTOL.com: From your perspective, what are some of the things that really differ when you’re doing manned test flights [versus unmanned ones]?
JP: Well, clearly the risk is a lot higher. You don’t want to kill a pilot — particularly I don’t want to kill a pilot! — so there’s a big shift in terms of the level of preparation, safety mitigation and analysis. You can run unpiloted flight and carry some business risk; if you lose the airplane you just lose some carbon fiber and electrics. If you lose the pilot obviously it’s a very different thing. The way you approach flight test is the same, it’s just your level of preparation and safety assurance has to be much, much higher.
eVTOL.com: So I know that [eVTOL developers] are taking very different approaches [to when in the development of their aircraft they start manned test flights]. Where on that spectrum does Joby fall?
JP: Well, we’re looking obviously for a manned operational vehicle, piloted, because I don’t think that either the regulators or the public are ready to get into an unpiloted autonomous vehicle yet. Clearly the industry is going that way, and we want to go that way, and I’m sure it will be here sooner than any of us think, but the initial approach is to operate our vehicle as conventional aircraft in the national airspace system, following current rules. So it will be fully certified, with a fully qualified commercial pilot, and our flight test program is aiming toward that.
Our flight test vehicles are optionally piloted, which gives us a great mix of being able to fly high-risk points unpiloted. So we can handle efficiently some high-risk test points, and prove the vehicle and expand the envelope if need be, and then put a pilot where he’s needed to close the circle — find out is this aircraft handling as we expect it to, pilot in the loop. So you’ve always got to put a pilot in at the end, but by being optionally piloted you have the ability to be very intelligent in your selection of what you do piloted and what you do unpiloted.
eVTOL.com: Can you say how you envision [the S4’s] mode of operation, in terms of the type of takeoff you do, or the type of landing you do?
JP: We envision a vertical takeoff and landing the vast majority of the time. We have the capability for a conventional takeoff and landing, or a rolling takeoff and landing; and semi-jet borne as we call it in the Harrier and the F-35, or semi-thrust borne, anywhere in the spectrum between zero knots and conventional takeoff speed. The energy saving from a conventional takeoff is less than you’d imagine, because with our very simple flight controls, you get airborne, accelerate, and you’re on the wing in a matter of seconds.
eVTOL.com: In terms of the emergency procedures for this aircraft, are they basically going to be, “land now”?
JP: There will be “land ASAP,” which is “land now.” In some of our procedures there will be “land as soon as practical.” There will be some emergencies that might require us to do a conventional landing; there will be some emergencies, like a stuck gear, that will say you must do a vertical landing. Everything is according to what failure there is. But one of the beauties of VTOL, you know as a helicopter pilot, is the flexibility that a vertical takeoff and a vertical landing offers you to put the vehicle down almost anywhere you need to. And particularly with the simple flight controls we have, to be able to do that without a high level of pilot skill, in fact with a very low level of pilot skill — you can put it down pretty much anywhere.
eVTOL.com: What do you see [being required for] pilot training for this new generation of vehicles?
JP: Well, in the early days, Joby will be flying fully qualified, commercially rated pilots. You could look to an intervening step — before you pluck that pilot right out of that vehicle completely — where you’ve got an operator or a driver who is not trained to anything like the same level a commercial pilot is today. What that training will look like, who knows, but it certainly seems like a logical step that you could have a halfway house between a commercially rated pilot and full autonomy.
eVTOL.com: It seems to me that with public acceptance being such a big piece of this puzzle, having professional pilots at the beginning is going to be critical to that. But what I haven’t really seen from the industry is anything that’s going to make commercial pilots want to go fly these vehicles. Pilots are kind of being discussed as this necessary evil on the way to full automation. From your perspective, what’s cool about flying these vehicles? Why would people want to do this?
JP: Flying a VTOL aircraft is fantastic, as you know. I think people will want to. I think there are a couple of different sources for our pilots, and we’re going to need a lot of pilots as we grow, before we get to the autonomous stage. So one of those is, it’s a great route to the commercial airlines [as a way to] build hours. But I anticipate that a lot of people will want to keep doing it, because with flying for the airlines, there are pros and cons. Some people love it, some will do it for years, but some of my friends are just absolutely fed up with it, with time away [from home] being the principal thing. If you’re flying with Joby Aviation in Los Angeles, you get to fly home in the evening, you don’t have a commute because you grab a Joby to get home, and flying vertical is great fun. And you’ll be flying interesting people, flying through the city, soaring over the traffic.
You know, there’s something that grabs the imagination about VTOL flight. I think that’s why this industry has suddenly come out of the starting blocks at such a pace. We’ve all sat in traffic and wished we had a vertical takeoff button in our car, and it’s a passionate thing — it’s more than just the dispassionate convenience of saving time. There’s something special about taking off like a butterfly and alighting like a bumblebee, and soaring over all the traffic stuck below. And low-altitude flight as well; we’re not going to be stuck up at 30,000 feet like airliners where everything on the ground is a little dot if you’re lucky and there are no clouds. So there’s an emotional attraction to this industry which I think will keep pilots flying Joby airplanes, even when they might have other options.