Skip to main content

Carmakers make cars. Polysync makes them drive themselves

autonomous cars require self driving software and that may change the world polysync technologies 003
Henry Ford changed the way cars are designed and built when he started the first assembly line on December 1, 1913. Before he flipped that switch, cars were made one at a time, and each one was a little bit different from the one before. Henry changed the game: cars could suddenly be mass-produced and sold at a price where the average guy could afford to buy one.

Ford made that change, and everyone else followed, but Ford also started the habit of keeping everything about his cars strictly in-house. He even had farms raising cattle to supply leather for his seats and beef lunches to his workers. In the business world, that’s called being vertically integrated, and it worked well for about 100 years. As it happens, times are changing again, and new companies are challenging the notion that automakers have to re-invent the wheel every time a new technology comes along.

Making a difference

With a solid century of experience, automakers have gotten really good at mechanical processes like stamping sheet metal, casting engine blocks, and assembling cars, but most of the underlying technology — the stuff that makes your car work as designed when you turn keys and press buttons — is completely unremarkable. Think about it: when you picked out your last car, did you base any part of your decision on the engine management software or the embedded code in the stability control system? Probably not.

When you picked out your last car, did you base any part of your decision on who wrote the engine management software?

The truth is that most of the underlying tech in any car is not part of the car’s value proposition or appeal. So why do automakers spend billions on custom solutions that don’t make a difference in customer satisfaction? And more to the point, why are they doing it all again as they rush to bring self-driving cars to market?

“What I saw was that companies were solving the full vertical problem of autonomous cars – everything from the operating system to the communications layer,” entrepreneur Josh Hartung of PolySync Technologies told Digital Trends. “They were trying to do the whole thing themselves, the same way they do everything.”

Hartung has a different idea for automakers: why not use a standard base of proven reliable software, and concentrate development efforts on features that make a difference to customers?

“Our product isn’t a differentiator, it’s an accelerator,” Hartung said. “What we have is a software platform that makes it easier to build and deploy autonomous driving applications. Whether it’s computer vision systems, neural networks, or control systems; these are all built on top of our underlying infrastructure.”

A data center on wheels

That may sound like a small thing – coordinating the inputs of different sensing devices to help build a self-driving car – but like all revolutionary changes, the implications go far beyond the first application.

“We’re trying to bring the velocity of software development to the necessary validation of an automotive ecosystem,” explains Nathan Aschbacher, PolySync’s Chief Technology Officer.

The OSCC Project / Demonstration

Because we all care about safety, automobiles are among the most regulated products on the market. Automakers have developed a highly protective engineering culture that emphasizes detailed planning to meet government regulations and liability standards. This in turn means that to recoup big investments in custom engineering, automakers prefer long product life cycles, and that retards the pace of change.

“In the auto industry, they define their software about five years ahead of production, which is why it looks five years out of date when it comes to market,” Hartung says. “So what each automaker has done is build this bespoke network of different hardware devices and embedded software. They end up not being able to take advantage of the benefits of software, like fast iteration. What we’re doing is decoupling the hardware life cycle from the software.”

Getting old-line companies to abandon the formula that has delivered them a century of success isn’t easy. But as automakers struggle to apply resources where they’ll make a difference, Hartung points out that a new model for software development can reduce the cost of cars as well as enhance reliability.

“Most of the functionality of the car can be defined in the software system.”

“Our view of the future of cars looks a lot like a data center on wheels,” Hartung proposes. “Most of the functionality of the car can be defined in the software system, and then the sensors, actuators, screens, and buttons are commoditized peripherals rather than smart boxes. This is a lot more flexible and lower cost for the automakers, because it’s a simpler system. And it becomes more robust.”

An additional benefit to a standardized software platform is the ability to upgrade cars in the field without a service appointment.

“Undoubtedly, this will be one of the biggest changes we’ve ever lived through, in terms of technology that impacts our daily lives,” says Hartung. “You’re seeing the first signs of what that will look like in companies like Tesla. They’re able to ship cars without some functionality, and then roll in software-defined functionality later. What we offer to automakers is an infrastructure that is inherently protectable and modern. We know that cars will be connected, so it’s not a choice.”

Getting some recognition

The Los Angeles Auto Show is changing its identity to Automobility, recognizing that fundamental changes have already begun to shake the industry’s traditional models. As part of that transition to a mobility-based model rather than simply selling cars, the show instituted a competition for automotive startups last year. This year, PolySync was part of the contest.

Over 100 automotive startups competed for a spot at the show as one of the Top Ten Automotive Startups. Once at the show, the number was reduced to three finalists, and then one winner. Of course that winner was PolySync, or we’d be talking about some other company right now. The L.A. award came with some cash, but the real prize was an invitation to the 2017 Extreme Tech Challenge (XTC) sponsored by Sir Richard Branson at his Necker Island estate in the British Virgin Islands.


“We’re really excited to be involved in the competition. We care so much about the ecosystem of startups and seeing more people getting excited about autonomous automotive. Really, the industry wins, and the award is a great vote of confidence for our business model. I think it’s a great sign for the entire industry,” Hartung said.

Moving forward in the background

You may never hear about PolySync again, but that’s not because they’re gone. By the nature of what they do, they’re close to invisible. But when you buy your next car and notice that it now updates itself like your phone or your watch, PolySync will be there.

“We’re underlying technology. It’s not ours to hog the spotlight,” Hartung insists. “We love cars, and I don’t see the demise of cars any time soon. What I do see is an industry that’s struggling to keep up with technology. You’ve got autonomy and expectations of infotainment systems being so much better than they were before. In an era of electrification and autonomy and mobility-as-a-service business models, automakers can no longer do it all internally.”

Editors' Recommendations