Working On Electric Aircraft Safety

The rapid development of electric aircraft intended for use a air taxis, such as Volocopter raises challenges for air safety designers.

As Evan Ackerman puts it, “what keeps an eVTOL aloft when things go wrong?” [1]

In a ground vehicle, robot or piloted, there are obvious and simple “safe landings”.  Cut the power, come to a stop, everybody has a good chance to get out safe.  And conventional “large” aircraft and copters can glide at least a little, which provides a last ditch chance to survive a catastrophic problem.

But small eVTOL craft do not necessarily have any gliding ability.  A loss of power results in a quick drop to the ground, which is likely to be unsurvivable.   This doesn’t matter for small toys, but as the technology moves toward commercial cargo and even human passengers, this is a challenge for air safety regulators.

This is not to say that eVTOL are particularly unsafe.  But it is clear that conventional methods of assuring air safety do not fully apply to what is termed “distributed propulsion”—e.g., multiple independent rotors. 

As noted, Volocopter has 18 electrically powered rotors.  So there are 18 independent propulsion systems that can fail.  If these systems are truly independent, then failures should be independent, too.  And clearly, loss of a single rotor or even a few should be survivable.  But if a single failure leads to a cascade of events that causes multiple failures, i.e., there are hidden dependencies, then a single failure can be fatal.

Ackerman reports that Volocopter has gone to great lengths to try to make sure that the systems really will fail independently, through “redundancy and dissimilarity”.  Every system has a backup, and the backup is built of different components and software.  (It is interesting that one of the differences is to use a different programming language for main and backup. That’s not worrying at all. : – ( )  They also use different companies to produce and validate the primary and secondary.

“Volocopter’s approach to safety involves multiple layers of both redundancy and dissimilarity. Every critical system has a backup system, and each backup system uses a different kind of hardware running different software written in a different programming language, all produced and validated by different companies. This insulates the overall system against any individual point of failure. But what about dual or even triple failures?”

(From [1])

The idea, of course, is to try to make sure that the supposedly independent systems aren’t just identical copies which could all hit the same flaw at the same time.

But with so many systems, there are a lot of possible combinations of failures.  Trying to show that the aircraft has a one in a million hours chance of failure requires that all the separate pieces be far more reliable than that.

Ackerman notes that electric propulsion systems do have the strong advantage of simplicity compared to piston or jet engines.  There are fewer things that can go wrong, and there is a lot of flexibility to react to problems.  (We also like electric UAVs because they are emission free.)

This is one industry that isn’t consumed by the Silicon Valley disease of “move fast and break things”.  Developers know that safety really is paramount.  One bad accident could kill the whole industry.  See:  Hindenburg.

Still, I’m not climbing in an eVTOL taxi any time soon.  I know far too much about software to trust it to be error free.  The description of Volocopters “multiple layers of both redundancy and dissimilarity” reminds me of the design of the Space Shuttle and other space systems, systems for which there are many situations for which there is no “Plan B”. 

I’ll note that your life insurance almost certainly doesn’t cover riding into orbit, and that’s a strong hint.  The risks are unknown but extremely serious consequences are very possible. Way, way beyond the risks of everyday life.

So—what would your life insurance company say if you decided to ride a Volocopter or similar air taxi? 

---

1. Evan Ackerman, Air Taxis Are Safe—According to the Manufacturers, in IEEE Spectrum – Robotics, October 21, 2021. https://spectrum.ieee.org/air-taxis-are-safe-according-to-the-manufacturers