FORD

Company Using Artificial Intelligence to Solve Urban Driving Problems

Ford's transition from automaker to mobility company took another step forward in a small office space in downtown Ann Arbor this week. Instead of a new car or fancy self-driving tech update, Ford's big news was, basically, an AI-powered database.
Standing next to a big 3D model of the city, Ford's vice president of mobility, marketing and growth, Brett Wheatley, announced the Ford City Insights platform. It uses AI and data from various sources—among them traffic cameras, parking garages, and police reports—to analyze everything from where collisions are most likely to happen to which roads would be best served by microtransit shuttles or scooters.
The City Insights platform is made up of four main sectors: safety, parking, transit, and a 3D model that makes sense of the other three. The AI system uses its deep data sources to answer questions that Ford engineers or city officials have. Are there enough parking spaces in town? Turns out there are, but most drivers don't know where the open spaces are. Discovering this fact meant that Ann Arbor decided not to build a new parking structure and instead will figure out how to direct drivers to open parking spaces.
What role do alleys play in the movement of a city? Ford set up sensors in a local alley and collected data on who used the space (more pedestrians than suspected) and which vehicles use it at what times (a delivery truck in an alley can cause traffic to back up in nearby streets). In a city where 85,000 people commute on the average weekday, understanding exactly who moves where when is an important first step in changing things for the better.


© Charlotte Smith/Ford Ford Mobility

The fourth part of City Insights is the City Insights Studio. It's made up of 3D-printed buildings and city blocks that sit on top of flat-screen TVs showing maps and data. The point of it is to better visualize the output from the other segments. All four of the parts that make up City Insights depend on the traffic cameras, parking garages, police reports on where collisions happen, and other rich data sources. Thanks to owners of Ford vehicles who have opted in to data sharing, the City Insights AI can also guess information including where accidents almost happen but don't—emergency braking incidents, for example—which is something that cities cannot know otherwise.
"At Ford, we're doing many more things than building vehicles," Wheatley said. "We're working to solve mobility challenges by partnering with cities."
Ford has been working with Ann Arbor for the past 18 months to get City Insights ready, and the automaker announced that it will expand its City Insights platform testing to six other cities: Austin, Indianapolis, Miami, Pittsburgh, and two other cities in Michigan: Detroit and Grand Rapids. Ford has worked with some of these cities on its City:One challenge mobility program. Ford said this week it is talking with other cities about City Insights but did not name them. Ford also hasn't said what it might charge cities to use the City Insights platform.
Back in 2015, Ford was already talking about "mobility" rather than simply concentrating on cars. By early 2018, Ford president and chief executive officer Jim Hackett was talking about a "mobility operating system." Wheatley said that City Insights is the evolution of what Hackett envisioned back then, a system where everything works together seamlessly.

Autonews