2014-02-26

Co.Labs

How GM Got Religion And Released An API

GM is betting big that software developers will help them take cars into the 21st century.



For much of the 20th century the automobile was American's sweetheart lifestyle product. In the last 10 years though, driving rates—especially among teens—have been on the decline. One of the main factors in that decline, according to a recent report by the U.S. Public Interest Research Group, is that driving isn't the cool, fun, hip pastime it once was. It's become drudgery.

Smartphones have replaced cars as the object of our collective identities. But now GM thinks cars can regain the "cool" mantle—or at least share it—by getting into the software business.

If You Can't Beat The Smartphone, Join It

GM wants its cars to be part of a platform. So last month it announced it's launching an API and app store which it believes will be home to a whole new segment of the 21st century software business.

I asked GM’s Greg Ross if it bothered him that cars can't fall back on the cool factor anymore. Ross is the director of product strategy and infotainment for General Motors' Global Connected Consumer group, and one of the people at GM driving the push to bring the car into the 21st century.

"Sure, it bugs me in that any time you're working on a product you want to make sure it’s as appealing as it can be to as many people as it can be," Ross says, "and youth more so than anybody because that's the future. I think we need to take it seriously to say, ‘What do we need to do to make car and car ownership more appealing?’"

Appealing is the right world. Apple and other smartphone manufacturers have "appealing" down to a science, which has helped fuel the boom in the mobile economy of the last seven years. Virtually every quarter there are smartphones with better screens, faster processors, thinner designs, and more addictive apps.

Contrast that with automobiles, which evolve only by degrees each decade, even as they get more expensive, and you can see why phones are more fun.

Of course, car manufactures are slow to "revolutionize" automobiles every year, unlike technology companies can with smartphones, because they’re obligated to adhere to more legal and safety regulations than a smartphone maker ever would be," Ross says.

"We're not on a cell phone development cycle, which make things, inherently, a little bit slower. But the car is on a long development cycle versus a smartphone and for good reasons. You've got several thousand pounds of steel going down the road, you have a lot of testing and development that needs to go into making sure everything works the way it's supposed to."

Developing For Machines That Travel At 70 MPH

This summer most of GM’s 2015 model year vehicles are going to have an AppShop application in their in-dash computers, GM's equivalent to the iTunes Store or Google Play. GM first introduced its API at the Consumer Electronics Show in 2013 and within the first year they’ve had a few thousand developers sign up, eager to work on an entirely new software platform.

However, while Ross and his team have seen many great app ideas from developers, they quickly found that it was important to work with developers to change the way they thought about how users interact with their apps.

"The developers we've run into have a lot of experience developing for phones and so we have to introduce them to the nuances and the unique requirements of developing for a car and to make sure that they're aware of things like driver distraction requirements and are aware of how we test our software," Ross says, and notes that because of safety and usability requirements the GM AppShop will only open with around 10 apps at launch.

While it’s unlikely anyone can point to an instance when operating an app has killed an iPhone user—and as alien as that possibility may seem—for developers who are going to delve into the new world of the car as a software platform, it’s actually something they need to think about.

"We test apps using the same standards we would use to test any in-vehicle executions," Ross says. "So we're driving around our proving grounds, testing for how long you need the clients away from the driving task while you interact with an individual app; to what extent does it interfere with the driving task, and applying the same standards we would to our own stuff. That's been a learning experience between us and developers."

Matter of fact, the very notion that you don’t want a user to be distracted by your app probably seems antithetical to most developers, which is why GM has lengthy guidelines on meeting distracted driving requirements for an app if it wants to be approved.

"Frankly, it's new to most developers and so we're having to do a lot of testing, a lot of validation, and a lot of education about what's doable," Ross says.

"We want to be on the right side of this. We're not introducing apps in order to create a distraction, we're trying to introduce applications, if anything, that are less distracting and more usable and formatted for car use than the alternative, which, unfortunately, right now is often customers looking down at their smartphones. We think designing an app framework that's explicitly formatted for use by the driver in the car is a way to do that and, frankly, that also makes us have to reject some applications because we can't come up with a way to do it in an appropriate manner in a car. That's been some of the lessons learned we're going through."

For The "Car As Software Platform" Developers Are Key

If there’s such a huge learning curve and GM has only approved about 10 of the submitted apps so far, I ask Ross why the company doesn’t just forgo the API and SDK and instead make its own apps.

Ross says the answer comes down to creativity. The iPhone, after all, was an amazing device when it was released in 2007, yet its full potential wasn’t realized until Apple opened it up to developers with the App Store a year later. GM believes the same great potential for the car as a software platform can only be realized with a vibrant developer community.

"There's a lot of creativity out there that we don't have a unique monopoly on," Ross says, and notes that developers will pick up on a driver’s needs just as good or better than GM might since each driver is unique. "There's other parties that have interest in developing applications unique to their own interest and values, so they're more willing to develop a niche application for their use than we might be."

"Take an example like, say, a particular home security company who wants to make the car able to interact with their whole home security system. General Motors isn't likely to make an app to work with a particular home security company, but if we have an application environment that allows something like that we can add value to our cars and rely on the motivating party to make the application that works in that environment. That satisfies the customer, satisfies that partner, and makes our cars a little more valuable, so I think that's just an example of what will allow us to have a greater diversity of content, tap into additional creativity, and really write off all the lessons learned that have been established in other industries."

We’re Looking For A Few Good, Relevant Apps

Speaking of other industries, Ross says GM is under no delusion that it is competing against app stores like Apple’s or Google’s for download numbers or the number of apps available.

"I wouldn't draw out a number but I don't see us having millions of apps," Ross says. "I don't think this is an iPhone platform. We want to get applications that have a meaningful value to your car ownership experience."

In other words, GM isn’t trying to make your car into the next great casual gaming device, so developers interested in making the next Flappy Bird probably need not apply.

"The initial discussion about apps in the car all focused on how to repeat popular smartphone apps in the car," Ross says. "There’s certainly a place for that to the extent you can take a popular smartphone app and format it better and more appropriate for car use—and I think that's a good thing. But, what we most want to do is: We're in the car business, we want to make better cars so we're looking for apps that make your car a better car."

Porting a streaming music service to the car? Sure, it makes a car more entertaining, but GM wants developers to look at the whole spectrum of vehicle ownership and operation and ask themselves how they can use the new API and SDK, combined with GM’s rollout of 4G in the car later this year, to make the entire automotive experience better.

What kind of apps? With a platform, Ross says, "I think I can make maintenance more convenient; I think I can reduce your insurance cost; I think I can reduce your fuel cost by giving you guidance on how to drive more efficiently along a more efficient route; and so on and so forth. Everything from shopping for the car, to initially getting oriented, to using the car, to maintaining the car, to insuring the car, to fueling the car, to using the car to do your day-to-day tasks. That's what we think would enhance the usefulness of the car and those are the kinds of apps that we're most in pursuit of."

The Road To The Future

Will GM’s API initiative change the auto industry in the same ways the iPhone changed the mobile phone industry? After speaking to Ross it’s clear that the company is betting it will. But that doesn’t mean it won’t take a while.

"This is a new model for our industry and so there's a lot of challenges just doing anything new like it. Everything is currently built around this notion of the car is designed to a set of specifications, tested to those specifications, and then released to those specifications." Breaking that 100-year-old mold, Ross says, is "a pretty significant shift for us as a company and for us as an industry."

However, thanks to GM’s lead in trying to make the car the next big software platform, it’s an industry that might now be on the upswing—and just may be enough to make cars "cool" again for the next generation of teenagers.

[Image: Flickr user Jim Larrison]






Add New Comment

2 Comments

  • Kurt Osterbusch

    Their strategy seems pretty shallow at this point - basically designing apps for a car connected device (instead of a phone connected device), and designing them to be "driver" friendly.

    It seems that real success might be found if the apps are directly tied into features of the vehicle itself, and those features would be unlocked by buying the app to control that feature. It may be a superfluous example, especially for GM (Scion maybe?), but a car could come with programmable LED lighting, that would only be controlled by an app. A transmission tuning app would be really cool as well. Each example requires a deeper dive into the overall design of the car, and not just the computer in the center console.

  • This could be revolutionary, however GM's closed minded attitude about it will probably make this a flop just like most other innovation put forth by GM over the last few decades.

    The blueprint for the car of the future exists: the Model S. Not to be a Tesla fanboy but they are literally the only company to have done anything revolutionary in this marketplace since... the 1950s?

    GM needs to recognize the simple fact that its "users" (drivers) are probably driving with the steering wheel in one hand and their smartphone in the other. How can we integrate these two beyond the simple bluetooth connection?

    Give me screen mirroring onto a built in TFT display. Give me control over what GPS app I'd like to use in the car. Give me the ability to drive my car from my phone, like James Bond in GoldenEye.

    Nothing close to this will come to fruition because as always, GM is trying to be on the cutting edge of innovation without taking any risks. No risk, no reward. Sorry GM.