Getting to Pluto with Self-Driving Vehicles

Ask any rocket scientist: The best way to get a spacecraft to Pluto is not to point your rocket at Pluto and light off the engine. No, you should instead point your rocket at Jupiter, where you can achieve an extra gravitational boost. That boost will give you velocity that your rocket could not achieve otherwise, and you’ll reach Pluto before any rocket sent directly towards it.

In the self-driving vehicle world, Pluto is the equivalent to “robo-taxis in large cities.” This is the market to which virtually every self-driving company aspires. The market is huge — as many of Lyft and Uber riders can attest — and self-driving technology will transform players in this space from massive money losers into even more massive winners.

But like Pluto, robo-taxis commercially operating in large cities is a long way off. In the meantime, we will continue to see massively subsidized tests and pilot programs where well-funded companies demonstrate that their technology is “almost there,” but those demonstrations are far from a viable product. Based on the current rate of technological progress, it could be well over a decade before we reach Pluto — err, I mean commercially-viable robo-taxis.

If Jupiter gets rockets to Pluto faster, what gets self-driving vehicles to the robo-taxi market faster? I believe it’s low-speed shuttles on fixed routes. This market has vastly reduced technical complexity: at lower speeds, sensors don’t need to “see” as far; the vehicle’s stopping distance is greatly reduced; and fixed routes mean that the AI must master only a limited set of roadways.

The shuttle market has other advantages over other potential “Jupiters” like retirement communities and suburban deployments. In both of those use-cases, the ride demand density is low: there may be 50,000 residents in a single retirement community, but to service them, the vehicle must master a huge network of roads. Worse, the actual rider demand is unknown: if users like a company’s service, the potential growth in demand could easily outstrip the company’s ability to deploy vehicles into the market. That creates a “success disaster” — where initial rider success fuels greater demand which cannot be met, which triggers longer wait times, and ultimately destroys the riders’ initial satisfaction. The shuttle market is protected from this effect because the ridership on many routes is at least approximately known.

May Mobility is going to Pluto, but we’re getting there via Jupiter. And we’re bringing actual riders on our vehicle with us. We’ve turned on a priceless flow of data — not just technical data, but also rider acceptance data, and operational (fleet management) experience. This will help us grow our company, our technology, and our customer base faster than other self-driving players.

Waymo and other big players may have launched their rockets before us, but they aimed directly at Pluto. If they get to Pluto, they’ll find the May Mobility flag already there.