In testimony before a Senate Subcommittee,Paul Misener, Amazon’s vice president for Global Public Policy noted that “While the FAA was considering our applications for testing, we innovated so rapidly that the [drone] approved last week by the FAA has become obsolete. We don’t test it anymore. We’ve moved on to more advanced designs that we already are testing abroad.”
The FAA took one and half years to give Amazon permission to fly one very specific model of drone. Misener compared this to what is happening overseas. “Nowhere outside of the United States have we been required to wait more than one or two months to begin testing, and permission has been granted for operating a category of UAS, giving us room to experiment and rapidly perfect designs without being required to continually obtain new approvals for specific UAS vehicles.
If it wasn’t so sad it would be too funny. Fun to watch an 800-pound gorilla throw its weight around – glad it’s on our side! There are some very fundamental differences between aircraft and drones – especially when the drone development is being driven by a company like Amazon. The speed of iteration (development) in the tech world enabled by today’s tools, plus the small size of the craft itself is radically different from what it takes to build a traditional aluminum and rivets airplane that can carry living, breathing human beings.
When you can print a craft on a 3D printer instead of having to first build jigs and tools, it is pretty clear that new ways of doing things are here. And that an enormous amount of time has just been slashed from development. In aviation after an airplane it is certified as safe and flightworthy, it is put into service along with a highly developed maintenance protocol. In tech, a solution is put into service warts and all, with the expectation that it will be improved in service.
There is a fundamental disconnect around safety. Logic suggests that testing the heck out of it before you let it out the door is essential when crew and passenger lives are at risk. But in this case, there are no crews and passengers, and the testing is being done in an isolated and theoretically safe area.
Why not iterate as fast as you can?
After all, no one at Amazon ever thought that their first design would be their last, or ever imagined that they would have to live with an interim one. Clearly at some point, Amazon will finalize a design because they need to in order to scale. Till then, we need policies that are built on the expectation of rapid iteration.