Three Philosophies, One Sidestick
Three Philosophies, One Sidestick: How Airbus, Boeing, and eVTOL Aircraft Actually Fly
Ask a pilot who has flown both an Airbus A320 and a Boeing 787 what the real difference is, and you'll rarely hear about engines, avionics suites, or cabin layout. You'll hear about the control laws. Both aircraft are fly-by-wire, meaning the pilot's stick or yoke doesn't physically pull cables to move a control surface anymore. It sends an electrical signal to a computer, which decides what the aircraft should actually do. That last clause, what the computer decides to do with your input, is where Airbus and Boeing diverge so sharply that pilots moving between the two fleets go through entire training modules just to unlearn each other's instincts. And now a third philosophy is emerging entirely, born not from decades of transport category tradition but from the particular demands of tilting rotors and distributed electric propulsion in the new wave of eVTOL air taxis.
What "Fly-By-Wire" Actually Replaces
Before diving into who does what differently, it helps to be clear about what fly-by-wire eliminated. In a traditional mechanically controlled aircraft, moving the yoke physically deflects the elevator through a system of cables, pulleys, and hydraulic boost. The pilot feels the aircraft's aerodynamic loads directly through that mechanical linkage. Push too hard at too high a speed and you feel the airframe fighting back, because you're wired straight into it.
Fly-by-wire severs that direct connection. A stick or yoke deflection becomes a digital input, interpreted by flight control computers, which then command actuators to move the surfaces. This unlocks enormous advantages: lighter airframes since you're not routing cable runs throughout the fuselage, more precise and repeatable control responses, and, critically, the ability to insert software logic between what the pilot asks for and what the aircraft is allowed to do. That inserted logic is called a control law, and it's the single biggest philosophical fork between Airbus and Boeing.
Airbus: Protect the Envelope, Command a Response
Airbus pioneered commercial fly-by-wire with the A320 in 1988, and its foundational design decision was to build the flight control laws around the concept of protected flight. In Airbus's default operating mode, called Normal Law, the sidestick doesn't command a control surface deflection at all. It commands a rate of pitch or roll, and the flight control computers figure out whatever elevator or aileron movement is needed to deliver that rate, continuously adjusting for airspeed, altitude, and configuration along the way.
Layered on top of that rate command is a set of flight envelope protections. Pull the sidestick all the way back in Normal Law and the aircraft won't exceed its maximum angle of attack, full stop. It simply won't stall, at least not through the pilot's own input. There's no need to manage a stall warning by easing off back pressure, because the computer has already capped how far back pressure can push the nose up. Similarly, bank angle is limited, load factor is limited, and if the pilot releases the sidestick entirely, the aircraft holds level flight rather than continuing whatever pitch or roll rate was last commanded. This is the origin of the famous Airbus philosophy: the aircraft assumes the pilot always wants a stable, protected outcome unless told otherwise, and the sidestick is a request rather than a direct command.
Crucially, Airbus's Normal Law isn't the only mode. If enough sensors or computers fail, the system degrades gracefully through Alternate Law, where some protections drop away, down to Direct Law, where stick input maps straight to a proportional surface deflection much like a conventional aircraft, and in the most extreme case, Mechanical Backup, a literal set of manual trim wheels and rudder pedals wired straight to the tail for a small number of Airbus types. The philosophy is explicit: give the computer maximum authority when it has full information, and hand authority back to the pilot in clearly defined stages as that information degrades.
Boeing: Feel Like an Airplane, Even With a Computer in the Loop
Boeing's fly-by-wire debut came later, with the 777 in 1995, and the company took a deliberately different stance. Rather than building a system where the stick commands an abstract rate and the aircraft protects itself absolutely, Boeing designed its control laws to preserve the tactile feel and control philosophy that generations of pilots had trained on in conventional aircraft.
The most visible manifestation of this is the yoke itself. Boeing kept the traditional control column rather than adopting a sidestick, and more importantly, it built in artificial force feedback. Push against the column and you feel resistance that increases with control surface load, mimicking the aerodynamic feel a pilot would get from a purely mechanical system. On the 777 and 787, that yoke is also linked electronically so that if one pilot moves their column, the other one moves in sync, replicating a mechanical interconnect that never actually exists in the wiring. It's synthetic feel, engineered deliberately to keep the aircraft feeling familiar.
Boeing's envelope protections are also philosophically softer. Rather than hard limits the pilot physically cannot exceed, Boeing built what are often described as soft protections, where the system provides increasing resistance and cueing as the aircraft approaches a limit like stall angle of attack or bank angle, but a pilot can still push through that resistance with sufficient force if the situation genuinely demands it. The idea is that ultimate authority remains with the human in the loop, with the computer acting as an advisor and a very strong nudge rather than an absolute gatekeeper. This isn't a minor detail. It reflects two different answers to the same design question: should the machine ever override the pilot's judgment entirely, or should it always leave a path for human authority to win out, even in extremis?
Neither philosophy is simply "safer" in the abstract. Airbus's hard protections have prevented loss-of-control accidents by making certain pilot errors structurally impossible. Boeing's soft protections have allowed pilots to make control inputs beyond normal limits in genuine edge cases where the computer's assumptions about the situation didn't match reality. Each approach has also been implicated, directly or by association, in accident investigations that shaped subsequent design revisions on both sides. This is a live, ongoing conversation in flight control engineering, not a solved problem with one obviously correct answer.
eVTOL Aircraft: Control Laws Built for a Different Kind of Flying Entirely
Everything above describes control laws built for aircraft with a fixed relationship between control surface and aerodynamic effect: an elevator always changes pitch, a rudder always changes yaw. eVTOL aircraft like Joby's S4 and Archer's Midnight break that assumption completely, and it forces an entirely new category of control law.
These aircraft typically use six or more independently controlled rotors, some fixed, some tilting between a vertical, helicopter-like orientation for takeoff and landing and a forward, propeller-like orientation for wing-borne cruise flight. In hover, there's no airflow over conventional control surfaces to generate aerodynamic force, so pitch, roll, and yaw have to come entirely from differential thrust across the rotor array, conceptually similar to how a multirotor drone stabilizes itself, but scaled up to carry passengers and certified to transport category safety standards. In cruise, the aircraft behaves more like a conventional fixed-wing airplane, with wings generating lift and, depending on the design, some conventional control surfaces coming back into play.
The flight control computer on an eVTOL therefore has to manage a continuous transition between two entirely different control allocation schemes, deciding in real time how much of a pilot's commanded pitch or roll rate should come from rotor thrust differential versus aerodynamic surfaces, all while individual rotors are physically rotating between vertical and horizontal orientations. This is a substantially harder control allocation problem than anything Airbus or Boeing's Normal Law or synthetic feel systems ever had to solve, because the number of available control effectors, the rotors, their tilt angles, and any aerodynamic surfaces, vastly exceeds the number of things the pilot is actually trying to command, namely pitch rate, roll rate, yaw rate, and vertical or horizontal acceleration. Distributing that authority intelligently across a dozen or more actuators, while continuously reallocating it as the aircraft transitions phases of flight, is precisely the kind of problem that makes eVTOL flight control software so much more computationally intensive than a traditional airliner's.
Redundancy philosophy looks different too. Airbus and Boeing achieve fault tolerance mainly through multiple independent flight control computers cross-checking each other, alongside triple or quadruple redundant hydraulic systems. eVTOL designs generally can't rely on hydraulics for a comparable safety case, so redundancy instead comes from the propulsion architecture itself: distributing lift and control authority across enough independent electric motors and rotors that losing any single one, or even several, still leaves enough authority for controlled flight and landing. That's part of why the FAA's conformity reviews for aircraft like Joby's S4 have specifically scrutinized fly-by-wire redundancy and propulsion system reliability as among the most technically contested elements of the entire certification case, alongside the broader question of what certification basis even applies. Because no existing FAA category, airplane or rotorcraft, fully captures a tilting-rotor eVTOL, Joby and Archer are both certifying under a bespoke pathway that lets the agency assemble certification requirements specific to the vehicle rather than forcing it into rules written decades before anyone conceived of distributed electric propulsion.
There's also a philosophical throughline worth noting. Where Airbus emphasizes hard envelope protection and Boeing emphasizes preserved pilot authority with soft cueing, most eVTOL programs are being designed from the outset with an eye toward eventual reduced-crew or fully autonomous operation. That shapes control law philosophy from day one in a way retrofitted onto a 1988 or 1995 airliner design never had to contend with: the control laws aren't just mediating between a human pilot's intent and the airframe's physical response, they're being built as a foundation that can, eventually, operate with progressively less human input in the loop at all.
The Same Question, Answered Three Different Ways
Strip away the specifics and every one of these systems is answering the same underlying question: when a human gives a flight control system an input, how much should the machine trust that input, and how much should it substitute its own judgment about what's actually safe? Airbus answers with hard limits and computed protection. Boeing answers with preserved authority and synthetic feel. eVTOL manufacturers are answering with a control allocation problem so complex that the very notion of a single, static "control law" barely applies, replaced instead by continuously adaptive software managing a transition between two entirely different modes of flight.
None of these are finished conversations. Every fatal accident investigation involving fly-by-wire systems feeds back into how the next generation of control laws gets written, and the eVTOL industry is currently having its own version of the debates Airbus and Boeing had thirty and forty years ago, except compressed into a few years of flight testing rather than decades of fleet experience. Given how much is still being worked out at the certification stage for the current crop of air taxis, this is very much a live and evolving area of aerospace engineering, well worth checking manufacturer and FAA updates for the latest, rather than treating anything about eVTOL control law ma
Comments
Post a Comment