To me, the most important thing isn’t speed or predictability. It’s price. Current delivery methods require a human being. People are expensive. If a delivery drone removes the human being from the equation then that could remove a significant fraction of the price.
Why couldn’t land-based delivery vehicle become autonomous though? That would also cut out the human in the loop.
One reason might be that autonomous flying drone are easier to realize. It is true that air is an easier environment to navigate than ground, but landing and taking off at the destination could involve very diverse and unpredictable situations. You might run into the same long tail problem as self-driving cars, especially since a drone that can lift several kilos has dangerously powerful propellers.
Another problem is that flying vehicles in general are energy inefficient due to having to overcome gravity, and even more so at long distances (tyranny of the rocket equation). Of course you could use drones just for the last mile, but that’s an even smaller pool to squeeze value out of.
In general, delivery drones seem less well-suited for densely populated urban environments where landing spots are hard to come by and you only need a few individual trips to serve an entire apartment building. And that’s where most of the world will be living anyway.
To me, the most important thing isn’t speed or predictability. It’s price. Current delivery methods require a human being. People are expensive. If a delivery drone removes the human being from the equation then that could remove a significant fraction of the price.
There are multiple land based delievery methods that don’t require a human: https://www.gearbrain.com/autonomous-food-delivery-robots-2646365636.html
Why couldn’t land-based delivery vehicle become autonomous though? That would also cut out the human in the loop.
One reason might be that autonomous flying drone are easier to realize. It is true that air is an easier environment to navigate than ground, but landing and taking off at the destination could involve very diverse and unpredictable situations. You might run into the same long tail problem as self-driving cars, especially since a drone that can lift several kilos has dangerously powerful propellers.
Another problem is that flying vehicles in general are energy inefficient due to having to overcome gravity, and even more so at long distances (tyranny of the rocket equation). Of course you could use drones just for the last mile, but that’s an even smaller pool to squeeze value out of.
In general, delivery drones seem less well-suited for densely populated urban environments where landing spots are hard to come by and you only need a few individual trips to serve an entire apartment building. And that’s where most of the world will be living anyway.