Robots and the Last Mile Problem
3 December 2018
One of the technological changes that I see on the horizon is autonomous robots going from labs and factories to cities and streets. It’s going to change a lot about how we interact with the physical world, and will have far reaching implications in pretty much every sector of the economy. One sector is transportation and package delivery. I specifically want to focus on the last mile problem, which is one of the most famous and thorny problems in modern shipping and transportation. If robots are able to overcome all the challenges associated with mapping, production, safety, etc. then it would be as big and impactful as google maps.
One of the biggest tests of big data and neural networks would be to deliver packages to every person in the US. It would give whoever does it a wealth of real world data that could be very useful in developing future robots. There would also be massive scaling benefits of the technology, because as more people use robotic delivery, the delivery becomes better as more data is generated. Despite all the recent hype, this is will probably happen before self-driving cars, as robots delivering packages are unlikely to kill people, so there needs to be less testing of them. I also don’t think the drone delivery will take off (heh) for anything except light packages that need to be delivered quickly. People don’t usually like having to worry about things falling on them (which people will worry about even if they’re super reliable), and there’s also the problem of packages that are heavier than they can carry.
We can already see some of this occurring in electric scooters and skateboards. They are sort of a solution to the last mile problem with human transportation. The difference, however, is that ultimately they’re controlled by humans, not robots. Some of this is already being tried with food delivery, though it’s been legislated away by San Francisco. I’d be interested to see how this plays out and what these companies will do in the next few years.