Off the Kuff Rotating Header Image

Drive.ai

Driverless car pilot ends in Frisco

I’d love to know what we learned from this.

For about eight months, a fleet of bright orange self-driving vans have been rolling around Frisco. The autonomous vehicles — and the pilot run by Silicon Valley-based Drive.ai — will be permanently parked on Friday.

The city of Frisco announced today that the autonomous vehicle pilot is ending. When it launched in late July, it became the first self-driving car service on public roads in Texas. The approximately 10,000 people who work in Hall Park, a large office campus in the suburb, could request a free ride in an app. The vans drove them a short distance to nearby shops and restaurants.

Nearly 5,000 unique riders used the service during the pilot program, according to the city of Frisco.

Drive.ai continues to operate another autonomous vehicle pilot in Arlington. The free service, which is available Monday to Friday, is open to the public and available in the city’s entertainment district.

Frisco Mayor Jeff Cheney did not say why the city decided to end the service rather than expand it. The city is encouraging current riders “to explore and utilize other modes of transportation after the conclusion of the pilot program,” he said in a prepared statement.

See here and here for the background. Do we have any idea how many riders per day used this service? Five thousand “unique” riders sounds decent, but not if 4,900 of them only ever used the service once. There’s a lot of talk about driverless cars as an enhancement or competitor to mass transit. Well, transit is measured in ridership, so let’s hear how Drive.ai did on that score. There are also numerous similar pilots coming, in Texas and elsewhere. They sound great, but until we see some data we can’t begin to evaluate their potential. I sure hope Frisco and Arlington give us the goods.

Autonomous cars in Arlington

Who wants a robot to drive them to a Cowboys game?

Arlington visitors and residents will soon be able to request an autonomous vehicle on demand in the city’s entertainment district.

The city approved a one-year contract with Silicon Valley-based Drive.ai to offer a new way for people get to Texas Rangers and Dallas Cowboys games, attend concerts at the stadiums or go to restaurants or bars nearby. Arlington City Council approved the contract Tuesday.

The service will begin with a fleet of three autonomous vans on Oct. 19, according to a news release. Each van will hold three passengers. The vans will travel alongside other cars, but will be programmed to operate in a designated area. They will travel at up to 35 miles per hour.

Initially, each van will include a safety operator. The fleet may expand to five vans, if needed.

As the story notes, Drive.ai is also piloting a program in Frisco, where as it happens the Cowboys are headquartered. This kind of fixed-route, short-distance, low-speed use of autonomous cars makes sense to me, though if it’s ever going to be more than a novelty it will need to be done at a higher volume than this. Starting out like this is fine – I’m sure there will be plenty of refinements to make to the idea – but to make sense and be cost-effective and a means to reduce traffic you’re going to have to figure out how to move a lot more people at one time. We’ll see if Arlington is thinking along those lines.

Self-driving car service coming to North Texas

Coming to the city of Frisco in July, a public-private venture pairing up with a California-based outfit called Drive.ai, for something that’s a little like shuttle service and a little like mass transit on a small scale.

The initial service will be available to transport the 10,000 employees working at offices at Hall Park to retail and dining options nearby at The Star in Frisco, where the Dallas Cowboys are headquartered. For many, the distance (just shy of a mile) is too far to walk but too short to warrant a trip by car.

People will be able to request a ride through a smartphone app. The service will be free during a six-month test run. Negotiations are already under way to bring a more permanent service to Frisco after that.

[…]

Safety is a priority, Andrew Ng of Drive.ai said at the Frisco event. And that’s why working with local authorities is so important. The company will be able to coordinate with first responders, help with public awareness campaigns and offer routes that add value. Local officials will also coordinate with the company when there are special events or road closures that affect traffic flow.

“Together we can make this thing as safe as possible,” Ng said.

Artificial intelligence is great at maneuvering fixed routes but has difficulty recognizing hand signals from a construction worker directing traffic, Ng explained. That’s where local leaders can step in and help.

He asks people to be aware of the bright orange self-driving vehicles, be lawful and be considerate around them — just like drivers are when they see school buses on the road. The vehicles also have four external screens to communicate with pedestrians and other drivers on the road.

Initial trips will have a human available in the driver’s seat of the orange vehicles to take over at a moment’s notice. The next stage puts the person in the passenger seat as a chaperone to answer passenger questions. The final stage lets the self-driving cars go solo with a remote operator available if needed.

Drive.ai is shouldering all of the costs involved in the pilot project. A dollar amount is not being disclosed.

“We’re invested in the region,” said Conway Chen, vice president of business strategy for Drive.ai. “We see this as a great test ground for other cities.”

James Cline, president of the Denton County Transportation Authority, said he believes self-driving vehicles have a place in public transportation and mobility. Whether that role is transporting people on that last mile from a bus stop to their house or replacing buses entirely remains to be seen.

“The challenge is going to be getting people to accept it,” he said.

This Fortune story has a bit more about Drive.ai, which I’d never heard of before now, as well as a map of the rute this car will follow. It’s not to scale, but given the description in the story my guess is that if it were more pedestrian-friendly, maybe more people would walk instead of needing a ride. Or maybe I’m just projecting. If the idea here is to make transit more feasible in these non-pedestrian-friendly places by solving the last-mile problem, that seems like a good thing. If not, we’ll just have to see. For this arrangement, Drive.ai – more likely, the venture capitalists funding Drive.ai – are paying for everything. How this might work in the real world is another question I’d like to examine. We’ll check back later in the year. The Trib and Texas Monthly have more.