General Motors Relies on IoT to Anticipate Customers’ Needs

Steve Schwinke, a member of the original design team for General Motors’ OnStar service and director of its Global Connected Customer Experience unit, has this to say about how GM leverages the Internet of Things to deliver products and services that consistently ensure the safety and security of its customers: “When I think about the Internet of Things (IoT), I think of three things: you can extract information from a particular device; you can share that information with other devices so they can communicate with each other; and you can add new services and new features.”

While Schwinke notes that GM has been engaged in IoT engineering for nearly two decades, “we’ve gone from 2G connectivity to 4G LTE connectivity, so multiple devices are connecting with unheard-of speeds.” Schwinke spoke with Michael Fitzgerald, contributing editor at MIT Sloan Management Review.

It seems that the work you’ve been doing at GM was the Internet of Things before we knew it was the Internet of Things. When did you realize that?

I’ve realized it gradually over the last few years. When I think about the Internet of Things (IoT), I think of three things: you can extract information from a particular device; you can share that information with other devices so they can communicate with each other; and you can add new services and new features. We’ve been doing all those things at GM really for 20 years. The big change is that we’ve gone from 2G connectivity to 4G LTE connectivity, so multiple devices are connecting with unheard-of speeds.

Is what you’re doing now with OnStar different from what you were doing when you started? Is it different because of the change in the underlying technology?

The 4G commitment has had a profound impact on our ability to expand what we can do. One of our proof points of 4G is vehicle health management (VHM), which is one of OnStar’s many services. People may not understand the level of sophistication and the amount of data that you need to look at the health of [their] vehicle — to see if there’s a potential issue, and then to alert them that they may have a potential problem. To do that, you need to extract large amounts of data recursively in order to look at the performance of your system and to look for abnormalities. Then you determine if that abnormality is something that could cause a problem in the future.

Can you give me an example of a vehicle health management initiative that you’ve launched based on your 4G capability?

Right now, we’re looking at starting systems. We want to make sure that you never have to walk home because of a dead battery. Starting systems are pretty robust across OEMs [original equipment manufacturers], including GM, but we still continue to monitor them and look for abnormalities and alert the customer when they have a problem. It’s a problem that most OEMs are challenged by. OnStar is introducing more and more electronics, so we need to make sure that the batteries in our vehicles are properly maintained so they can recharge the battery and the whole starting system whenever they need to.

General Electric has been beating the drum to use IoT to become a services company. Are you seeing opportunities through the kinds of warning systems you’re talking about to create new services that might create revenue streams?

I think we’re always exploring ways to create more value for GM, but we’re also using the data to reduce our costs. For example, we can extract a lot of data from the VHM service to help us understand how our vehicles are performing in the field and give us better insights to the types of improvements we need to make.

Is GM thinking about connectivity as a way to improve driving habits? I’m particularly thinking about the driving habits of teenagers.

GM is very involved with driver safety for teens, but I like to think that connectivity will help everyone be a better driver. We actually have a new service coming out called OnStar Smart Driver that is designed for customers who want to know more about their driving habits and gives them feedback on how to be a better driver. Customers can then go online to get info on their specific driving habits, like hard braking and hard acceleration, their overall Smart Driver score, and to see how they compare with a group of aggregated OnStar customers also enrolled in the program.

We’re following the safe-driving guidelines that the OEMs adhere to, but I think we can use connectivity to profoundly affect the way everyone drives. That would be a difference maker for us going forward.

Talk to me about that process of managing your Internet of Things. What’s the management thinking that goes into how to use this technology effectively?

My big thing is, do we have the right enablers in place, the right hooks in place so that future customers as well as our current customers can take advantage of the IoT extension? As we start looking at more and more devices that can be connected while you’re in the vehicle, what are the enablers we need to put in place to create value-added service?

What are some of the most challenging non-engineering issues that you face when you’re developing these new services and figuring out new ways of connectivity?

I always talk to my team about the Wayne Gretzky quote — skate to where the puck is going. How good are we at really anticipating? What are the things that our customers need but don’t know they need? I want to invent things that they don’t know they need. It’s figuring out where the industry is going.

How do you get people to think that way? Do you send people to hang out with the tech visionaries at Silicon Valley or bring in the futurist of the month to talk about trends?

Actually, it’s a combination of things. We send people to the Google and Apple conferences every year so they see what’s going on in the industry, so we’re tapping into the heartbeat of what’s going on in the Valley. We’re doing the same thing with the people we acquired when we bought Cruise [Automation]. We have a team in San Francisco because we bought a company called Sidecar, and now those developers report to me. And we have teams across the globe. I also have an innovation forum in Israel and in Germany as well as here. We’re tapping into the best talent around the world to understand what’s coming next.

Acquiring the Sidecar team has been really eye-opening for me. We both bring value to the equation and we’re thinking in a constructive way, which makes us a winning combination. The synergy between our two organizations is compounding daily in terms of what we can accomplish together. But it’s not easy.

Say a little more about that because it’s a classic conundrum. GM is one of the largest companies in the world, and the classic problem for a large company that buys a small company is that the cultures are so different you don’t get the value you expected. As a manager, how are you dealing with that?

You make sure you understand both sides. You’ve got to listen, and you’ve got to explain. The way you’re going to be successful is by understanding how the other company operates. Sidecar is a company with a very strong brand and a very strong commitment to their customers. We have to respect that and not be dismissive of it. Otherwise the acquisition will fail.

How do you design this Internet of Things technology, with its rapid development and rapid innovation, so that it fits into the traditional automakers’ three- to five-year cycle?

I think General Motors actually recognizes that there are these two entities, two immovable objects coming at each other. How do you stay current with consumer electronics when you have customers replacing their phones every nine months and also have a vehicle that has a 10-year life cycle and a three-year development process? How do you fit those two things together? How do you continue to evolve the software when it has to change in sync with the release of the vehicle? It’s a challenge, but we’re committed to getting it done.

You’ve been at GM for 19 years, through the whole evolution of OnStar. How has GM managed the transition to being more data-oriented?

Safety and security is what OnStar is known for, and it’s still at the forefront of what we do. However, we’ve been providing customers with access to vehicle data for a number of years through OnStar Vehicle Diagnostics. Now we’re just beginning to tap into the potential of the much-larger data pipeline enable by 4G LTE. Accessing and interpreting this data is incredibly beneficial, whether those benefits come through as new services, like Proactive Alerts, or behind-the-scenes benefits as we determine the best parts for our vehicles.

To gather the data, to extract the data, is not trivial. We have done a lot of that in house. In fact, the software for latest version of the subsystem that does our data collection was actually developed in-house. We acted almost like a tier two to a tier one in order to put that software in the module.

GM also buys a lot of components, and the OnStar module is one of them. And this is like a multiple generation of how to pull data out of the vehicle. We just started doing it ourselves because we were the closest guys to it. You’re not going to write a [stack] to encapsulate all of the things that we’ve learned over the years, for the past 10 years, trying to do data-driven services. So we are really proud of the functionality in one of our modules that allows us to actually have so much access to information.

But ultimately, you had to develop expertise and build that technical knowledge in-house?

Yes, and a lot of that has to do with continuity. There is a kind of ebb and flow to how things are designed and developed, and you don’t ever want to lose that continuity of expertise, so you need to manage it carefully. If you can’t manage it, it’s really hard to control. We’ve developed some really strong core competencies here that allow for that continuity.

And you keep learning. You don’t always need the new kid to come in. Sometimes it’s the old salty dogs that add a lot of value.

There’s a classic communication gap between the technical side and business side that companies are always struggling with. How has that played out at GM over the last 19 years?

In large organizations, walls create problems between business and engineering — and as they grow, the walls get bigger. More process gets introduced. But with smaller companies, where you don’t have all those restrictions in place, some really cool things happen. We’ve always tried to make sure those walls don’t get too high so that we can maintain those opportunities for constant dialogue.

One last question before we end. What’s the big challenge in bringing IoT into an existing organization?

The big challenge is having your leadership recognize the value in what you’re trying to do. If they do, they’re going to make the commitment to invest, and we’ve had a high-level commitment from our leadership to do that. I came here from a different OEM because of the commitment GM made 20 years ago to do this work. If you demonstrate a commitment to do something, people recognize it. People really want to take pride in the work they do. They want to be able to look at their neighbors, their friends, and say, “Hey, I did that.”

That’s how we stay competitive, too. We make a commitment to deliver new services and new offerings so people can talk to their friends about them.

In talking with you, it really feels like GM’s long-term vision is finally becoming a reality.

The longer-term visions are absolutely there, but we still have this DNA of commitment to making sure we deliver safety and security. We won’t ever forget that we’ll get help for you when you can’t get it for yourself.