Ford plans to build a fully autonomous vehicle without a steering wheel or pedals by 2021, the company announced today. The vehicle will be aimed at ride-hailing and ride-sharing fleets.
To get there, Ford is working with a number of startups, including a new investment in Velodyne, a firm that makes LIDAR sensors; the acquisition of SAIPS, an Israeli company that makes computer vision and machine learning software; an exclusive licensing agreement with “virtual retina” technology company Nirenberg Neuroscience, LLC; and a previously announced investment in 3D-mapping company Civil Maps.
The company is also expanding its research and innovation center in Palo Alto and plans to double the size of its team there by the end of next year.
Ford executives said the company wasn’t at the stage where it could show a vehicle, or even talk about what that vehicle might look like. Instead, it’s focused on getting the hardware and software ready for a driverless car.
The goal is to build a vehicle to the SAE Level 4 standard of automation. This would give the car the ability to handle all aspects of driving, although limited to certain approved areas or regions. A car could be restricted to only the island of Manhattan, or to take certain approved routes outside of Manhattan to get to the airport or Yankee Stadium.
“In a ride service, you could imagine that the defined environment or area might be large enough to take a customer from a city center to an airport or a seaport,” said Ken Washington, Ford’s VP of Research and Advanced Engineering. “Depending on how much of that environment can you capture in your high resolution map, you can define the area that you’re going to service with the vehicles.”
This makes the task a bit easier than building a car that can handle every possible situation, as vehicles will only be allowed to drive in areas that have very high-resolution mapping data already.
Over the next five years, in addition to working out the software and hardware, Ford and the rest of the industry will need to get governments to write regulations allowing them to offer these cars at all. Ford, Google, and Uber joined forces earlier this year to form an autonomous car lobbying group.
Ford plans to sell the cars initially to ride-sharing services, presumably like Uber and Lyft, though Ford demurred when I asked about any potential partners. It also declined to say if it was considering the launch of its own ride-hailing service.
“We are viewing this just like the core business of selling cars,” said Washington. “We stand behind our product. We envision a future ride service having vehicles in it that we engineer and service, and provide a service to our customers.”
It’s not clear if Ford thinks this particular technology — where there’s no wheel or pedals at all — will ever make it to consumer vehicles, or if it will only find a place in car fleets. But the company is not planning to sell it to consumers initially. Eventually, after several years as a ride service, the autonomous technology may be ready to sell to the public.
“It will be more expensive” than a traditional owned vehicle, thanks to the need for extensive computing power and sensors, said Washington. “The miles driven are likely to be much higher as well. When you put a vehicle that has the ability to be autonomous into a ride service, it will be designed to operate in a defined environment, managed by the right business who will amortize that asset over many more miles.”
In other words, expect this thing to be a lot more expensive than the venerable Ford Crown Victoria that filled the country’s taxi fleets for decades. Ford views the fleet environment as one that is “more amenable” to a first introduction of this technology, thanks to the cost and mapping requirements, but it expects the costs to fall over time and that it will be more feasible to sell to consumers down the line.
The competition is out there, though — Google has had a similar car (meant to have no wheel or pedals) in testing for several years. Uber is spending huge sums on its own autonomous car efforts too, including poaching a number of engineers from an autonomous driving lab at Carnegie Mellon University.
Google and Uber seem to be going head-to-head in the race for self-driving taxis, and now Ford is joining the battle. Though of the three, only one of them has any experience actually building cars. Whether that gives Ford enough of an edge remains to be seen, but the company is confident in both its plan and its timeline.
After years of seeming to be just around the corner, the era of self-driving taxis could be just five years away.