Self-driving is not a killer feature — it’s a game changer.
Suppose you’re at a car dealership in a few years’ time. The salesperson is pitching a self-driving upgrade option. “Hmm,” you think: it’s an expensive upgrade; it’ll only work sometimes; my car is still a “car” (and not a mobile office or movie theater), and I’ll still have to endure traffic delays and search for parking.
You may even contemplate that you, with your single-occupancy vehicle, would be contributing to the problem of city congestion, and while you may be able to pay the $10+ congestion fee to go south of 60th street in NYC, that same fee decreases access and equity for the majority of the city’s population.
“Nah,” you say to the dealer, “I guess I’m just not that excited about self-driving cars.”
In that instant, you join the ranks of 78% of Americans who say they wouldn’t pay more than $2,000 for a self-driving car feature¹. Self-driving simply isn’t a killer feature — it might be a “nice-to-have”, but it doesn’t fundamentally make your transportation better.
Let’s step back for a moment.
New technologies are often viewed through the lens of existing products and services — take an existing product, add a new feature, then sell it.
In 1994, a new technology emerged that reduced the storage requirements of digital music: the mp3. (Mini-refresher: an mp3 is a type of file that digitally stores audio — it uses data compression to make those files very small.) Existing music-player products quickly sprouted mp3 playing capability, allowing them to store about ten times as much music as before. But it was basically the same product and customer value proposition — you buy the music and you put it on the device. These products didn’t make discovering new music any easier, and consumers had to organize and manage their albums that could contain thousands of files.
Those first products did benefit from the new technology, but they missed the transformative potential of mp3s. The impact of mp3s ultimately wasn’t that you could fit more songs on your memory card, it was that it enabled streaming and online music services, which made enjoying and discovering new music effortless. The ability to play mp3s wasn’t a killer feature, but streaming music services (which were made possible by mp3s) transformed the industry. This benefited not only consumers, but also created some of the most enviable businesses in tech, such as Spotify.
Today, most autonomous vehicle companies are bolting self-driving capabilities on top of existing products. While a taxi or personally-owned-vehicle that can drive itself has some advantages over its traditional counterpart, the lasting impact of self-driving vehicles is something completely different.
To understand why, we need to understand what’s actually wrong with transportation. Consider this:
- Buses in NYC average 7.1mph — they’re barely faster than walking.
- Uber and Lyft cars, once viewed hopefully as a solution, actually make things worse: the average occupancy for a ride-hail vehicle is between 0.3–0.5 — in other words, for every person actually going somewhere in a Lyft, there are one or two more vehicles circling for their next fare and adding to congestion.
- Personally-owned vehicles must park. Parking structures are expensive (multi-level structures can cost more than $30,000 per spot to build), and use up real-estate that could be better used for green space, business use, or housing. In New York City, for example, the land dedicated exclusively to parking (ignoring parking-within-buildings!) covers more than 2.2 times the area of central park².
Transit systems like buses and shuttles — the things that could alleviate congestion — often provide a poor rider experience today. This poor experience arises largely from wait time and inconvenience: the transit systems don’t pick up and drop off when or where riders want, they take indirect routes, and they stop frequently³. Low satisfaction pushes riders to use other solutions, including greater use of personally-owned vehicles and ride-hailing services, both of which make congestion even worse.
Imagine a city where the transit system is convenient, fast, clean, and reliable… in short, a transit system that’s so pleasant that people use it. Wherever you are, there’s a comfortable car only a minute or two away, and it’ll take you within a minute’s walk of your destination. If that transit system could average only 3.4 people per vehicle⁴, it would reduce traffic by half. Traffic would flow easily through the city, allowing higher speeds and a better ride quality, further increasing rider satisfaction.
Building a system like this requires a large number of low occupancy (around 6 people) vehicles. Instead of a big bus coming by every 20 minutes, you have smaller vehicles coming by every 2 minutes. Low vehicle occupancy is also necessary to provide direct routing for riders — otherwise, you back to a local “bus” which needs to stop frequently. And the system needs to be centrally managed and coordinated — it has to be good at matching riders to keep occupancy up, it has to be good at finding pick-up and drop-off locations that don’t disrupt other travelers, and it needs to be integrated with traffic control systems to keep traffic flowing efficiently. It also needs to coordinate with the mass transit systems that are important to many cities — trains and buses are great solutions when you have hundreds (or thousands!) of people going in the same direction at once. The rider experience must be seamless when changing between a bus and a lower occupancy vehicle. That requires a choreographed dance between different modalities so that efficiency, convenience, comfort, and safety are achieved simultaneously.
Here’s the problem: this kind of system cannot be built with human-driven vehicles. The cost of the drivers is a major obstacle, but so is the limited availability of those drivers⁵ in the first place. If we can’t solve the problem with drivers… well, I think you know where I’m going with this.
Self-driving will save cities from congestion — not by being deployed in personally-owned vehicles for the affluent, but by fundamentally enabling high levels of service for riders in public transportation systems. High satisfaction will drive adoption, allowing our cities to become more accessible, vibrant, economically prosperous, and equitable for everyone who lives there.
May Mobility’s vision is to transform cities using our self-driving technology, making them safer, greener, and more vibrant. Our customers generally do not care that our cars can drive themselves. But they care tremendously that we can provide a level of service that leads people to choose to take our shuttles rather than driving themselves or taking an Uber. For our customers, self-driving technology isn’t a killer feature, but it is the key enabler for transportation systems that work.
Today, we’re live in Detroit, Columbus, Providence, and Grand Rapids. These deployments have not solved those cities transportation problem… they’re still small in scale. But we have proven that we can create tremendously satisfied riders, and we have measured the changes in our riders’ habits.
Come ride with us! Our Columbus, Providence, and Grand Rapids deployments are open to the general public, no reservations required.
- https://www.insurancejournal.com/news/national/2019/04/02/522530.htm
- http://whatthestreet.moovellab.com/
