With U-M Tech, Ford Research Car Maps Its Surroundings in 3D
A press preview event put on last week ahead of January’s North American International Auto Show had, as one might expect, the relentlessly positive feel of a pep rally. Joe Hinrichs, Ford’s president of North American operations, told the crowd that 2014 “represents a massive year for Ford.”
Less than a week after the Dec. 12, of course, Ford (NYSE: F) would warn that 2014 will not be as profitable as 2013, sending its stock price tumbling. But Hinrichs did have some genuinely good news to share, including Ford’s plans to launch 16 new products in North America in 2014 and add production at seven North American plants, creating 5,000 jobs.
But for tech geeks like me, the real excitement began when Raj Nair, the vice president of product development, took the stage and was joined by Ford’s automated Fusion Hybrid research vehicle, which was developed with help from researchers from the University of Michigan’s College of Engineering and safety data from State Farm Insurance.
Nair was careful to explain what we were seeing, a white Fusion with four whirring canisters mounted on its roof, wasn’t a driverless car, but rather a car that can operate automatically as long as a driver is inside. The research car was equipped with driver assistance technology that incorporates cameras, radar, and algorithms to allow the vehicle to handle stopping and starting on its own, and has the ability to correct driver errors.
The Ford research vehicle also utilizes LiDar sensors, which use light the way a bat uses sound waves. The thermal images created by the sensors, combined with images from the 360-degree mounted camera, can generate a 3D map that detects moving objects.
A live feed showing the 3D map from the research car’s dashboard was projected onto a big screen at the press conference; what we saw was an image of the conference room and the car surrounded by reporters, all rendered in technicolor waves (click on the image above to see it in all its trippy glory). Even this cynical reporter has to admit it was pretty cool.
U-M played a big role in developing the algorithms used to generate the map. Alec Gallimore, associate dean of research with U-M’s College of Engineering, called Ford “an exceptional partner.” Ford and U-M, along with MIT, have been collaborating as part of DARPA’s automated vehicle challenge since 2007. Professors Ryan Eustice and Ed Olson spent many hours in university parking lots refining software that helps the research car see and navigate through the world, Gallimore said in a follow-up interview this week.
Gallimore also said the partnership with Ford has meant that U-M researchers have an avenue to see the immediate benefits of their work. “We’re really trying to tackle hard problems that address societal needs,” he added. “Car accidents, loss of life, injuries, the environmental impact of stop-start technology—[our research] says that perhaps there must be a better way to make the driving experience safer.”
At the press conference, Nair said the kind of automation on display in Ford’s research car is still years away from being on the market. In the next five years, however, he said we should expect to see cars that can help drivers park and navigate through traffic jams.