The future of in-car computing technology

At the South by Southwest Interactive festival, a Ford systems engineer opines on what consumers may be able to expect in terms of hardware/software platforms in their cars five years from now.

Daniel Terdiman Former Senior Writer / News
Daniel Terdiman is a senior writer at CNET News covering Twitter, Net culture, and everything in between.
Daniel Terdiman
6 min read

AUSTIN, Texas--I have seen the future of computing technology in cars, and it's not coming any time soon.

It is coming, though, and when it arrives, it may very well change the way we deal with information while we're driving. But because the auto industry moves at a truly snail's pace when it comes to innovation, it's likely to be at least five years before this vision comes to pass.

My view into this future came from a conversation I had at the South by Southwest Interactive (SXSWi) festival with T.J. Giuli, a vehicle software systems engineer in Ford's Infotronics Research and Advanced Engineering division.

A Ford screenshot showing the current state of the art when it comes to in-car technology. Over the next five years, though, hardware and software are likely to be integrated, making a wide variety of cloud- and sensor-based applications possible. But safety must still come first. Ford

Giuli began our conversation with a quick recap of where Ford has taken in-car technology over the last few years, concluding with the fourth, and latest, version of its Sync platform, which is allowing partner companies and developers to access Sync APIs and integrate some smartphone applications with in-dash displays. The idea here, Giuli said, has been to "become more of a part of the human/machine interface."

"But that's production land, and I'm research land," Giuli joked, and in his neck of the Ford woods, it's all about looking five years down the road. And that's where a proposed system he called Fiestaware comes into play. The idea of the conceptual technology would be to begin on the hardware side with a "big, beefy, desktop-class computer" embedded in the vehicles control systems and to link that to a 10.2-inch screen in the vehicle, as well as to 3G mobile broadband connectivity.

On the software side, he continued, the idea would be to integrate that computing power with a wide range of social-networking sites, like Facebook, Twitter, Foursquare, and others, and to find ways to build applications around those sites using data constantly streaming in from the car itself. As Giuli put it, "vehicles are moving sensor platforms...[and we] can also do interesting things if we have applications running that take advantage of sensor values."

There are already many applications available for devices like the iPhone that mimic some of the information such an embedded system might offer, Giuli said, like an iPhone app that uses that device's accelerometer and GPS to provide 0 to 60 miles an hour timing. But he pointed out that any such estimates are suspect because they are prone to a lot of error due to the sensors being extra to the vehicle.

But if the application and the sensors are both integrated with the car, he suggested, the information could be much more precise.

Another likely direction such systems could take would be marrying voice-recognition technology with various in-car applications, allowing drivers to have much more control with just their voice.

Why platforms matter
To Giuli, the biggest advantage of building hardware and software systems into cars is that it creates an integrated platform that could help the auto industry move a lot faster when it comes to providing drivers with the latest information-based technology.

He pointed out that though in-car navigation systems have been around for quite some time, they have been essentially modules with one function only and yet still required three-year development cycles. That means, he continued, that the Garmins and TomToms of the world are able to move much faster and far surpass what's possible with such in-car systems. Still, he said, to many drivers, having a built-in car navigation system is worth a lot, and so the car companies keep on including them, even though drivers could easily get more advanced technology from third-party manufacturers.

Nevertheless, he acknowledged that the auto industry has struggled to keep up with the consumer electronics design cycle, and that that dynamic is what is spurring development, at Ford, at least, of systems that will allow new software applications to be added as they are developed, rather than have drivers be stuck with the systems they got when they bought their vehicle.

"There's definitely a case for the notion," Giuli said, "that customers want to be able to change. The same user interface isn't necessarily appropriate for everyone."

A need for safety
On the other hand, as recent developments in the auto industry have demonstrated, safety has to be a car company's number one concern when it comes to new systems, and that is likely to mean that even if a car has a powerful computer built into it, the manufacturer is still going to need to maintain a substantial level of control over what kinds of applications drivers have access to. No car company's lawyers, in other words, are going to sign off on systems that would allow drivers to be able to play World of Warcraft while they're behind the wheel.

That's an exaggeration, of course, but the point is that Ford, and other manufacturers that adopt systems like the ones Giuli was talking about, will need to ensure that any applications available to drivers don't impact safety.

What that means, then, is some sort of managed openness, he said, a platform that third-party developers can be successful on, but that also gives the Fords of the world the ability to ensure that whatever applications are written are "driving-appropriate."

One likely scenario, Giuli said, is that data from the cars could be sent, via the 3G network, to the cloud, allowing for a series of more complex applications that could collect, analyze, and manipulate that data, and present it to drivers on their home computers. In that case, he added, the manufacturers' hold on the reins could be loosened somewhat, since they wouldn't need to worry so much about safety.

And while much of these kinds of systems might be five years away, Giuli said, some elements, such as a potential app store and the exposing of APIs for the fourth-generation Sync, could be just one-to-two years off.

A lot of inertia
Still, Giuli said, implementing the heart of this hardware/software vision is likely to be a five-year process, and asked why, he said, there are a lot of technological hurdles that, in some sense, require "changing a very large manufacturing company with a lot of inertia. Ford is a very technology-focused company, but also, in many ways, there is a lot of inertia."

But what's really five years off is not any specific genre of applications but rather the underpinnings of the platform, the notion of figuring out how the code actually runs on the platform, and of determining what the appropriate model is for user interfaces on the platform, as well as what kinds of data to expose to developers.

Giuli said the most likely base for any such systems would be Microsoft's Robotic Studio, which offers interchangeable software modules or services, and a service-oriented architecture. "You can basically add to that platform very, very rapidly," he said, "so it's a very, very fluid thing."

And while it might be nice for a company like Ford to think it can dominate a new in-car technology genre, Giuli acknowledged that it would probably be better for everyone concerned if the entire auto industry worked together on a common platform.

Obviously, he said, each company would want to be able to differentiate its offerings. "But just based on market size, it would be best if [the car makers] got together and thought about common platform features. That may be a ways off for organizational reasons...but if this is going to be a real serious effort, that's probably the way it's going to go."

Correction: This story initially misspelled the name of the Ford research program.