Ariel Moore exhaled sharply and lifted her arms to the sky. “I have arrived alive!” she said to no one in particular. “I have arrived alive!”
This should not be notable. Moore just took a half-mile ride in a six-seat shuttle, one of several that run in a loop between her office in downtown Detroit and the garage where she parks her car. But on that sunny June day, she and her colleagues at real estate company Bedrock also did something quietly remarkable. They become the first American customers to take a trip on a self-driving shuttle—for a fee.
Moore’s relief at her survival was understandable. She had heard something about a self-driving Uber killed an Arizona woman in March, and that Tesla’s semi-autonomous Autopilot feature has been linked to road deaths. Meanwhile, the engineers, the venture capitalists, the professional prognosticators, the auto executives, have gone a bit wavy on autonomy in the past year. The common refrain has gone from “It’s just around the corner!” to “This tech isn’t quite there.”
In front of the Bedrock parking garage, though, Moore’s colleagues mostly shrugged at the robot ride. One woman who tends to get car sick said she appreciated the smooth ride. Another complained the shuttle’s front door didn’t shut properly. Others went straight into the parking structure. It was 5:45 pm on a summer Wednesday. Let someone else be in the history books. The desk jockeys wanted to get home.
The WIRED Guide to Self-Driving Cars
For May Mobility, the ambitious autonomous vehicle developer behind the shuttle service, such boredom is the goal. In just two years, the Ann Arbor-based company has built the software that allows its vehicles to drive themselves, and arguably beaten the biggies.
Waymo says it will roll out a robo-taxi service in the Phoenix suburbs by the end of the year. General Motors’ Cruise says it will launch a ride-hailing service in 2019. Ford is testing pizza delivery in Miami, with plans for a bigger rollout in 2021.
May Mobility already has a roster of paying customers. There’s Bedrock, in Detroit, and a pending deployment in Grand Rapids, Michigan. The Ohio Department of Transportation is overseeing the launch this week of a 1.5-mile shuttle service in downtown Columbus. Also this week, the Rhode Island DOT is announcing a contract with the startup to provide a shuttle service.
“Our sales pitch is not that we are autonomous,” says Edwin Olson, the startup’s CEO and a veteran of the Toyota Research Institute's AV program. “It’s that we provide a better level of service and we’re solving real transportation problems.”
Problems like going back and forth between Bedrock’s parking garage and its office. Or complementing a three-mile segment of public bus route in Grand Rapids. Or providing a connection between two busy urban neighborhoods in Rhode Island. Or giving locals and visitors a more interesting way to travel through downtown, as in Columbus. All along fixed routes, all with a friendly attendant in the front row to help new or confused riders, and take the wheel if the tech falters.
“Our ideal partners are people that have first mile, last mile challenges, who are trying to help individuals get from things like transit stops or parking structures to their end destination,” says May Mobility COO Alisyn Malek, who first heard of Olson’s approach while working in venture capital at General Motors. 1 “What this gives us, essentially, is a captive audience, and a known road network, and nodes that people need to move between.”
That self-driving challenge is way easier than demanding a car can go anywhere, anytime. May Mobility doesn’t have to keep a meticulously updated and curated map of an entire city. And it dodges the Uber problem, the need to develop software that can efficiently match riders and rides and routes.
Yes, these limitations make the whole driverless thing much less sexy. No, this is not the self-driving future you have been awaiting. But it’s the one that’s here now, and May Mobility has good reason to be creating it.
Market research firm IBIS World estimates the “scheduled and chartered bus service” industry took in $5 billion last year. Malek says May could aim to target the half of US trips that are 3 miles or less. A handful of other companies have flocked to this opportunity: Optimus Prime operates a service on Boston’s Seaport, and within a planned community outside the city. France’s Navya just wrapped up a free, year-long shuttle demonstration in Las Vegas. Since October, Drive.ai has been running three vehicles along a few routes in Arlington, Texas. Still other shuttles are testing in even more controlled environments, like college campuses or dedicated testing grounds.
Which means companies like May might make some money—and garner it plenty of experience with what humans want out of a self-driving car—while they wait for the technology to catch up with that greatest promise, the go-anywhere-anytime robo-car. And that is the long-term goal, Olson makes clear. On one wall in his company’s office, the team has hung three banners, which you might call targets. They read Waymo, Uber, and Cruise.
Like every other city on the planet, Detroit is occasionally messy. When I rode in a May vehicle this summer, it had to stop—smoothly—for a toddler-toting man who stepped into the middle of the street. When a cyclist shot out of the Jimmy John’s sandwich shop against traffic, the May shuttle made a wide sweep around him. (“We see that guy all the time,” one employee told me.) A few times, the human operator took the shuttle’s T-bar steering wheel and did the driving: once because it needed to cross a double-yellow line to swerve around a parked car, and once after the vehicle braked hard for a yellow light.
That messiness is one reason May says it’s testing in busy, complicated urban centers, and not, as Olson puts it, “just a sleepy suburb of Phoenix.” (That’s a dig at Waymo.) And yet, that safety operator’s presence makes something clear: May is cheating, just a bit.
It plans to keep safety drivers—or attendants—in its vehicles for an indefinite period, to keep riders safe and informed about how the technology works. Its geographic reach is measly. In Detroit, five vehicles run that single one-mile route, which has been mapped and re-mapped to make sure the vehicles can pull it off. The company’s Rhode Island deployment, which will happen in stages starting early next year, will be its longest route yet, around 5 miles. The vehicles are slow, too, with top speeds of 25 mph.
And May uses what others in the industry see as a faux pas: sensors that live outside the shuttle, which the company mounts on traffic lights. These constantly send info, like street light color, to its vehicles, helping guide them on their path. Other autonomous vehicle developers say such help shouldn’t be necessary. “From an infrastructure standpoint we really don’t need a lot,” Waymo CEO John Krafcik told a group of governors in July. “We can’t become reliant on it at Waymo because sometimes it’s not going to work.”
May thinks such pretensions are bunk. “Let’s use what works in robotics today and go solve transportation problems, because that will allow us to learn more, quickly, and to scale more quickly,” says Malek. What she means is: Let’s do what we can now, learn from it, and make some moolah along the way. Documents provided by Ohio’s DOT show that May won its one-year Columbus with a $547,750 bid, $150,000 less than its closest competitor. Rhode Island’s government will kick in at least $800,000 for its one-year contract with May. (Some of that money comes from a federal transit pot.) These are not huge checks, but they’re something.
Still, the road is less than smooth for autonomous shuttle operators like May. In August, the Department of Transportation published a report on the sector, pointing to limited vehicle capabilities, opposition from labor, procurement complications, and the unpredictability of local regulations, politics, and funding as issues that low-speed, self-driving shuttles must conquer before becoming ubiquitous. “The market is small, and many companies in this space have little experience designing and validating systems and producing vehicles, compared to traditional automakers,” the report’s authors wrote.
For now, May must keep proving that it’s worth the money. In Detroit, Bedrock head of parking and mobility Kevin Bopp says the company is happy with the service, and is already considering new routes. “The feedback continues to be really positive,” Bopp says, noting May has provided 22,000 rides to employees thus far. Bedrock is also looking at opportunities to invite the wider Detroit public aboard the company’s vehicles.
In Columbus, the DOT initiative called DriveOhio that chose May Mobility’s shuttle proposal sees its run as a way to collect data on autonomous shuttles: whether they’re useful, what sorts of tech challenges they face right now, whether—and this is a biggie for governments—their procurement can slot into the limited and sometimes inflexible government process that exists today. “Our first shuttle, obviously, we’ll learn a lot, as far as any policy or regulations or requirements,” says Jim Barna, who heads up DriveOhio. “And then we’ll be able to use all of that for future deployments throughout the state.”
And so on Wednesday, the very first members of the Columbus public will get to ride May Mobility’s mobiles. And once they’re in the self-driving future, and hopefully, a little bored, they likely won’t mind if someone cheated a bit to get them there.
1 Correction appended 12/4/18, 11:10 AM EST: A previous version of this story misidentified Alisyn Malek at the CTO of May Mobility. She is the COO.
- What's the fastest 100 meter dash a human can run?
- Amazon wants you to code the AI brain for this little car
- Spotify's year-end ads highlight the weird and wonderful
- Hate traffic? Curb your love for online shopping
- You can pry my air fryer out of my cold, greasy hands
- Looking for more? Sign up for our daily newsletter and never miss our latest and greatest stories