Cloud power: A screenshot shows experimental Ford software that can predict where a driver will travel, based on previous routes. Ford is working with Google to develop a system that will analyze data in real time.
As lead evangelist for Ford’s digital technology efforts, Prasad is responsible for making sure Ford does not repeat its Wingcast mistake and that it pursues a “democratic” approach to technology with what he calls an “Apple-like” fervor. When I visited Ford’s headquarters in Dearborn, Michigan, in December, executives wore jeans and spouted jargon like “open-source,” “UI,” and “services-oriented architecture.” And it goes beyond talk. Some of Ford’s latest patents, like one covering a system to check a driver’s health, are basically complex network diagrams in which it’s hard to spot an engine block or an axle.
Many of Ford’s ideas for the future of the automobile were on display in August, when the company unveiled Evos, a prototype of its “cloud-connected” car. The flashy red sedan is a designer’s mock-up—complete with flip-up gull-wing doors that will never make it to a production line. Inside, Ford imagined glowing 3-D displays for live music feeds and real-time traffic updates, and also sensors that could monitor the driver’s heart rate and alter the car’s performance accordingly.
“With Sync we empowered the driver. Our next leap is into empowering the vehicle. With the cloud we can certainly do that,” says Ryan McGee, a technical expert in Ford’s Vehicle Controls Architecture and Algorithm Design research group.
An image from a Ford patent application for monitoring a driver’s health.
McGee believes the computing power available on the Internet will allow cars to become smarter. Last spring, at the annual Google I/O conference in San Francisco, McGee announced a deal to use the search giant’s prediction algorithms—online software that analyzes large data sets to spot trends. Ford’s idea is to send data from your car to Google’s data centers, which would then predict where you are headed every time you key the ignition. Google might predict, say, that there’s a 59.24 percent chance you’re headed over to Bob’s house. A hybrid car might use a map of low-emission zones to determine when to switch to battery power as you drive. Or the algorithm could pick a fuel-efficient path with few hills, no rain, and the least traffic.
“Fuel optimization depends on the topography, traffic patterns, and how a customer drives their car,” says McGee. “The cloud will allow us to use these three data points that historically were not aligned in real time.”
McGee says Ford has already developed beta versions of its prediction software. And real-time weather and map programs are currently two of the top three kinds of apps used on smart phones, according to ComScore. However, don’t hold your breath waiting for the prediction technology to reach the dashboard; Detroit’s product cycles are still longer than those of Silicon Valley. A Ford spokesperson says the rollout of the prediction software in Ford cars is “in the four-to-eight-year range.”