Showing posts with label mutual. Show all posts
Showing posts with label mutual. Show all posts

Thursday, August 30, 2018

Ford and Argo AI achieve mutual respect

Engineers from Detroit and Silicon Valley are like oil and water, right? So what happens when they come together to develop self-driving vehicles for the masses? How does that work?

Indeed, the idea of an artificial intelligence expert helping to shape sheet metal seems about as likely as a vehicle engineer specifying the number of teraflops for a general computing system. But as Ford and software partner Argo AI prepare the automaker's first autonomous vehicle for 2021 deployment, such collaborations are becoming commonplace.

In an interview with Automotive News, Argo CEO Bryan Salesky and Ford's chief program engineer for autonomous vehicle development, Chris Brewer, waxed enthusiastic about the cross-pollination of Argo's high-tech capabilities and Ford's vehicle development know-how.

They say the challenge of building a vehicle that can drive itself and hold up to the high utilization rate Ford says is needed to make the technology profitable is creating mutual respect.

"It sounds like putting Argo's stuff on a car requires a unique process," says Salesky, 38. "But it's no different than balancing the 100 different trade-offs that a chief engineer has to manage around noise, vibration and durability. This is ... what [Ford engineers] have to do on typical vehicle programs."

Rather than "bolting technology onto an existing vehicle," which Salesky calls "an unnatural act," his robotics experts are involved in the ground-up development of Ford's first autonomous vehicle.

For his part, Brewer, 57, acknowledges unique challenges associated with developing autonomous vehicles. Matching the lengthy development time required for traditional vehicles with the rapidly changing sensor, computing and software technology Argo needs to stay on the cutting edge requires obsessive communication, he says.

!function(d, s) {var ip = d.createElement(s);ip.async = 1, s = d.getElementsByTagName(s)[0], ip.src = "//s.idio.co/ip.js", s.parentNode.insertBefore(ip, s)}(document, "script");$(function() { $idoWidget = $('#idio-article-recommendations-8'); $imageContainer = $idoWidget.prev().prev(); $imageEle = $imageContainer.children('img'); if ($imageEle && $imageEle.length > 0) { $imageContainer.insertAfter($idoWidget); }});

"It's been a really interesting dialogue, having this 'clock-matching' discussion," Brewer says. "Bryan will say, 'We still need to figure this system out,' and it becomes the job of the Ford side to figure out how do we buy more time for them to figure that out."

The key to this collaboration, both say, lies in avoiding the silos that can keep engineers with different specialties from having input in other areas of development.

"The whole approach is to not have this 'throw it over the wall' mentality," Brewer says. "It's this permeable membrane between the vehicle side and the technology side that allows us to bring a product to market and then scale it in a more affordable way."

This approach gives Salesky unprecedented input on the vehicle design itself.

"I can say, 'I think we'll have a more cost-effective program if we can just change the shape slightly,'?" he says. "I can eliminate a sensor and come up with a more cost-effective solution. That's powerful."

Meanwhile, Brewer has also had to consider factors such as the size, heat and power draw of the processors needed to drive Salesky's artificial intelligence, as well as the "pipes" that transfer huge amounts of data around the vehicle. These challenges have made the autonomous vehicle program a magnet for Ford's most talented engineers, Brewer says.

The very different cultures of high-tech companies and automakers have the potential to create culture clashes, Salesky says.

"In a big tech company, you tend to have all the trade-offs swayed toward the side of function over everything else," he says. "Can we get the software engineers as much power as possible?

"Well, it's never enough. They always want more. To be part of a vehicle program you're forced to look at the realities of not just performance at all costs, you are forced to look at affordability, durability, reliability and so on."

Salesky's career began working on Carnegie Mellon University's entry in the 2007 DARPA Urban Challenge and was he technical project manager at Google's self-driving car program. But while at Caterpillar he helped create autonomous drive systems for giant dump trucks working in mines, preparing him for the kind of collaboration that the Ford partnership requires.

"Most of my career has been making things with wheels and tracks autonomous," he says. "We didn't necessarily build the things with wheels or tracks. So I've always had to work with big companies that have their traditions in how they build these products. … The thing I've learned is that they're all really good engineers that want to build a great product."

It's a shock to some, Salesky says, when engineers used to locking in "the iron" years before production try to collaborate with people working with technology that changes in a matter of months. But, he says, the common language of engineering quickly cuts through any differences.

"Once you're all aligned on the mission, you get good engineers in the room and magic will happen," Salesky says, with a smile. "I've seen it time and time again."

ATTENTION COMMENTERS: Automotive News has monitored a significant increase in the number of personal attacks and abusive comments on our site. We encourage our readers to voice their opinions and argue their points. We expect disagreement. We do not expect our readers to turn on each other. We will be aggressively deleting all comments that personally attack another poster, or an article author, even if the comment is otherwise a well-argued observation. If we see repeated behavior, we will ban the commenter. Please help us maintain a civil level of discourse.


View the original article here