Is HTML5 the silver bullet for automotive app development?
Andrew Thompson and Pádraig McGarrigle report on the future of HTML5 in automotive infotainment systems and whether it will bring carmakers the diversity of apps they crave.
For more than a decade, the World Wide Web Consortium (W3C) has been at work on a set of coding standards intended to enable a single piece of code – be it for an app or a website – to display across a broad range of connected devices. And the automotive industry has been paying close attention, hoping this new set of standards, known as HTML5, would increase the number of apps available for the connected vehicle, or, at the very least, bring down development costs.
By the look of it, the W3C, which acts as the main international standards organization for the World Wide Web, is getting close. In December 2012, it published the complete definition of HTML5 specifications – an actual set of standards is expected before the end of this year. And the connected car industry is now part of the official process. The W3C’s Automotive and Web Platform Business Group launched in February 2013, and the first draft of an HTML5 automotive standard is expected in April.
What’s more, HTML5 is already in use by multiple carmakers in some shape or form. Mazda is the latest to adopt it. But, as Roger Lanctot, associate director, automotive multimedia & communications service, Strategy Analytics, points out, HTML5 “is only as useful and cost-saving as the programmers and the carmaker that put it to work are clever.“ "HTML5 is just plumbing," he says. "It is very nice plumbing, but it cannot make up for flawed business models and poor app architectures."
HTML5 high hopes
There are high hopes when it comes to this particular kind of plumbing. One of the big problems the automotive industry has had in recent years is a dearth of apps for the connected car. And a big part of the problem is that independent app developers are reluctant to spend the time and money on automotive apps as long as they have to fit each application to an individual car brand, or even model within that brand, and all they get in return is a few thousand downloads.
HTML5, with its device-agnostic approach, can remake app-making for the connected car, says Andy Gryc, senior automotive product marketing manager for QNX and co-chair of the automotive board for the W3C. “We see it as a big advantage,” he says, adding that HMTL5 has the potential to bring mobile app developers and the connected car ecosystem closer together.
“It enables application access in a way that they wouldn’t have before,” he says. “Larger companies don’t have a problem. Facebook or Pandora, for example, those guys can actually afford to do custom and specialized things for the car brands that they’re involved in. [But] the vast majority of applications that you could have access to, that could add some very cool value to the car, do not have that luxury.”
The Facebook issue
Not everyone is as upbeat. In December 2012, Facebook ditched its HTML5-based apps in response to sluggish performance. Mark Zuckerberg later called betting too much on HTML5, as opposed to native apps, “the biggest mistake we made as a company.” (The argument was and still is that a native app – an app created specifically for a device or an operating system – allows developers to unleash the full potential of the underlying platform.)
Alex Bratton, CEO of app developer and consulting firm Lextech, warns that reliance on HTML5 by OEMs and automotive app developers may be a mistake of similar magnitude. “HTML5 has its uses – for example, with content look-up,” he says, adding, however, that native apps have the potential to deliver a far superior consumer experience. Another argument for native apps is, Bratton says, that the current automotive trend is not in increasing on-board apps but rather in more mobile apps that connect to the on-board system.
Still, Gryc, who is something of an automotive HTML5 evangelist, believes that the problems that beset Facebook will not apply to the automotive. “The types of applications that we are looking at are not going to be reliant on flashy graphics, continually moving images, animations and a lot of things like that,” he says. “You get that mobile eye candy [factor] when you’re trying to engage with mobile apps. But the automotive space is not like that at all. Any concerns about responsiveness are a little misplaced because the automotive space doesn’t need those kinds of concerns. You don’t want to have a display that’s continually changing in the dash, showing new messages and doing all kinds of scrolling and stuff because it [amounts to] driver distraction.”
Gryc is speaking from experience. “We built a whole automotive HMI with it, so we actually know it’s extremely capable,” he says. “The vast majority of things that you want to do in an automotive setting are easily handled in HTML5.”
There are other concerns, to be sure.
One is that HTML5-based apps will only work with an internet connection. In reality, it is possible to build a solution that works without connectivity, at least in some basic functionality.
Another concern is that HTML5 will open the floodgates to any developer writing applications for the vehicle. Not to worry, Gryc says. “We’re not talking about having a single app store for every OEM,” he says. “So I think that once a developer actually creates an app, it is up to the individual OEMS to say these are the kinds of apps I want to let into my car, for the demographics of the people who are actually purchasing my vehicles. It does let them tailor that experience.”
Ultimately, no matter which solution a carmaker chooses, it “will consume time, money and manpower delivering and maintaining an app experience in the car,” Lanctot says. “The least expensive path is to enable asmartphone-based platform leaving the bulk of the software updating to the mobile device,” he says. “Can HTML5 be leveraged to enable this? Yes, it can and already is. But so can Qt and Android and all the rest.”
Gryc agrees that what matters the most is making an app environment that is at least somewhat similar to mobile. “It allows third parties to come over to an auto-centric approach at a lower porting cost,” he says. “The mobile developers won't have to start from scratch to build their auto app. That's true whether you build it into the head unit or deploy it through a connected smartphone. One may be simpler than the other, but it has to be auto-adapted either way.
"Since a lot of the automotive activity in mobile/consumer apps is eventually being paid for by the OEM, using a mobile framework is a definite win, not only from costvbut from time to market and availability. Picking a platform that is broadly supported in mobile – HTML5, of course, but also Android and Qt – makes perfect sense in reducing cost and broadening the offer.”
Andrew Thompson and Pádraig McGarrigle are regular contributors to TU.
For all the latest telematics trends, check out Telematics for Fleet Management Europe 2014 on March 12-13 in Amsterdam, The Netherlands, Content and Apps for Automotive Europe 2014 on April 8-9 in Munich, Germany, Insurance Telematics Europe 2014 on May 6-7 in London, Telematics India and South Asia 2014 on May 28-29 in Bangalore, India, Insurance Telematics Canada 2014 on May 28-29 in Toronto, Telematics Detroit 2014 on June 4-5 in Novi, Michigan, and Advanced Automotive Safety USA 2014 on July 8-9 in Novi, Michigan.
For exclusive telematics business analysis and insight, check out TU’s reports: Telematics Connectivity Strategies Report 2013, The Automotive HMI Report 2013, Insurance Telematics Report 2013 and Fleet & Asset Management Report 2012.