Skip to main content

Self-driving pods are slow, boring, and weird-looking — and that’s a good thing

Self-driving pods are slow, boring, and weird-looking — and that’s a good thing


Driverless pods, retirement communities, and grocery delivery are the early hallmarks of a huge paradigm shift

Share this story

It’s 2018, and you may be asking yourself, “Where are all the driverless cars?” Well, they’re already here, driving at extremely low speeds in tightly geofenced areas in over a dozen cities, in what is probably the most low-key, unassuming start to a global paradigm shift that we’ve ever seen.

We can all be forgiven for assuming that completely unencumbered driverless cars were right around the corner. After all, the purveyors of such technology lead us to believe as much, promising coast-to-coast demonstrations and selling us on the prospect of sleeping while driving. But the reality is much different: low-speed autonomous pods are currently operating all over the country, racking up an impressive number of trips while barely drawing any attention for it. As billion-dollar companies like Google, Uber, Ford, and GM scramble to perfect their high-profile robot taxi projects before they launch, smaller, more nimble startups are making progress in lower-stakes pilots that are operating right under our noses.

Small, low-stakes self-driving pilots are making progress right under our noses

These pilots can be broken down into two categories: driverless shuttle services and self-driving delivery projects. Both are operating at low speeds — vehicles that max out at 25 mph are typical — and along structured routes or within tightly geofenced areas. These services “can help meet the complex mobility needs of neighborhoods, campuses and business districts, especially by being traveling in mixed traffic — alongside pedestrians, cyclists, scooter riders and more — and facilitating connections to other modes of transit,” Kelley Coyner, a senior fellow at George Mason University and founder and CEO of Mobility e3, wrote in Axios this week.

On Monday, May Mobility celebrated its 10,000th trip in Detroit, where its fleet of autonomous, six-seater shuttles offer rides to Quicken Loans employees for free along a one-mile loop. It only took the Ann Arbor-based startup 75 days to hit that mark, a sign that slow and steady can sometimes win the self-driving race. As required by Michigan law, each shuttle currently has a human fleet attendant in the vehicle. And in a city defined by automobiles, the friendly green-and-white shuttles are being praised for offering a reliable alternative to car ownership.

It only took 75 days for one driverless shuttle to hit 10,000 trips

French startup Navya has been operating a similar-looking autonomous shuttle in Las Vegas for almost a year. The shuttle crosses eight intersections on its circular route, has transported thousands of passengers, and has logged only one crash (when a truck backed into it on its first day in operation). Meanwhile, Local Motors has its 3D-printed driverless shuttle Olli on the roads in Washington, DC, and it’s looking to expand in California and Arizona.

These shuttles operate along structured routes like a streetcar, but other ride-hailing startups are giving their AVs more breathing room — but not that much more., a Mountain View, California-based startup, has four bright orange minivans operating within a two-square-mile geofenced area outside of Dallas, Texas, and will soon be making trips to AT&T Stadium, the home of the Dallas Cowboys. Voyage provides an on-demand self-driving taxi service for residents of retirement communities in San Jose, California, and Orlando, Florida.

But why restrict yourself to just human cargo? Every week brings an announcement of a new self-driving delivery service. Last week, San Francisco-based startup Udelv announced it had signed a deal to supply Oklahoma City’s largest grocery chain with self-driving delivery vehicles. Previously, the company’s bright orange vehicles were delivering groceries for the high-end Draeger’s Market chain in the Bay Area city of San Mateo. Now, they’ll be used to deliver a wider selection of food for the more mass-market customers of Uptown Grocery, Buy For Less, and Smart Saver stores.

Driverless delivery services are popping up all over

Another self-driving startup, AutoX, recently announced it would begin delivering groceries in its small fleet of retrofitted Lincoln MKZ vehicles for customers in San Jose. And Nuro, a company founded in 2016 by two veterans of Google’s original self-driving team, is teaming up with supermarket giant Kroger to deliver groceries to residents in Arizona.

When self-driving cars first entered the public conscious seven or eight years ago, the assumption was that ride-hailing and taxi services were the most obvious application. Take a normal car, add cameras and high-tech sensors, stuff some GPUs in the trunk, remove the driver, and presto! Instant future. The predictions were predictably pie-in-the-sky. One study projected that autonomous vehicles will create a massive economic opportunity that will scale from $800 billion in 2035 to $7 trillion by 2050. These cars were just going to print money for companies like Google and Uber.

Image: ABC 15

But recently, much of that hype has collided with the reality that self-driving cars will take much longer to hit the road than previously thought. Public support for the technology has waned in the wake of a fatal crash involving a self-driving Uber vehicle last March. A bill that would allow for the sale and deployment of hundreds of thousands of self-driving vehicles is stalled in the US Senate, and it’s not expected to pass anytime soon. Many critics now say that self-driving cars are stuck in the “trough of disillusionment,” a reference to the “Hype Cycle” popularized by research firm Gartner.

Against this challenging backdrop, autonomous vehicle developers are moving aggressively to launch commercial ride-hailing services, in which customers would pay to take trips in robot taxis. Waymo has said it would kick off its first revenue-making service in Phoenix before the end of the year. GM is eyeing 2019 for its commercial operation, most likely in San Francisco. Ford is laying the groundwork for autonomous ride-hailing and delivery in Miami.

Are self-driving cars stuck in the “trough of disillusionment”?

But ride-hailing in dense, urban environments is immensely complicated, and even companies that are the furthest along in the technology, like Waymo and GM, have struggled to develop self-driving vehicles that can navigate tricky traffic scenarios safely and without annoying other drivers on the road. In a recent interview with Fast Company, GM CEO Mary Barra was asked whether the automaker was still on track to introduce a car without a steering wheel and pedals by 2019. Her answer was telling: “We’re still on track and hitting our milestones to be able to deploy a self-driving vehicle in a fleet situation, inside a geo-fenced ride-share environment, next year.”

Barra, like other executives, is trying to better manage expectations for the inevitable rollout of self-driving cars. Waymo CEO John Krafcik recently said it will be “longer than you think” until self-driving vehicles are everywhere.

So while we wait for the driverless revolution to fully manifest, these small, painfully slow shuttle pods are a positive sign of what’s to come. For decades, driving has been synonymous with high speeds on the open road, and the tens of thousands of traffic deaths that occur each year were dismissed as the byproduct of unfettered vehicular freedom.

If we’re serious about eliminating those deaths, then we need to psychologically prepare ourselves for urban transportation that’s slower, safer, more boring, and a lot more like public transportation. As cities increasingly look to these services to attract employers, hopefully they will incentivize transit agencies to make real improvements to our bus and rail systems. Because self-driving technology that doesn’t work in concert with public transportation isn’t a revolution worth fighting for.