Leadership Talks - Eyes on the Road
June 1, 2007 By: GPS World Staff GPS World
Chet Huber (CH), president of OnStar, a wholly owned subsidiary of General Motors that provides in-vehicle telematics to five million subscribers, spoke with GPS World on May 7.
Alan Cameron (AC): OnStar’s Turn-by-Turn service provides customized voice-guided directions. You have added an option for automated audio commands in addition to live operators. Why?
CH: We first launched, in September 1996, completely around the concept of a live advisor in the whole category of navigation and directions. Part of that was because it was consistent with our brand, a very human-oriented high-touch experience, and part was that the technology at that time was just not far enough advanced when we started that we had anything that we could really be proud of as an automated solution. We’ve always been about trying to execute our services in a low-cost way in the vehicle and keep the experience very easy to use. So that set us on a path that said we weren’t going to rely heavily on embedded nav devices and screen-based systems.
As the technology has progressed and as our experience has grown, we’ve cycled through seven generations of hardware technology in the 10 years we’ve been in business, which for an automotive company is well beyond unprecedented in terms of how fast we could spin the platforms.
What that allowed us to do in Gen 7, which just launched, is combine what we think is the best of both parts of what our brand can be. It retains this intuitive user interface and usability around a live advisor to determine the destination, but uses obviously the sophistication we can build into the technology today to enable the vehicle now to voice-direct you to your intended destination, triggered off the GPS location of the car.
We think it really enhances this service from where we had been in the past. All the research we’ve done around comparing this alternative with a kind of traditional screen-based approach showed that the majority of people that have a navigation need would prefer the simplicity and intuitive nature of the way we’re executing our service.
We had the idea for OnStar in the middle of 1995, and had our first working prototype in January 1996. The good news for us is that we had a really compressed development cycle that let us get into market quickly from the time that all of the pieces were able to be snapped together.
AC: That’s pretty fast, put alongside automakers’ design-in times.
CH: It started with very senior commitment at the top of the company to move in an unprecedented, nontraditional way. We saw very early, when we keyed up the opportunity to do something like OnStar, we saw the potential power of the service and the capability, and particularly the most important aspects, which had to do with helping someone in the case of a dramatic emergency need like a crash or another emergency situation.
But we also saw that if we approached it like a traditional vehicle technology, that we literally shouldn’t start. Because, at the end of the day, we would be way behind what would be possible or what would make sense to consumers, if we approached it the way traditional vehicle development technologies would be approached.
And so at the very senior levels of the company there was this decision that said these services are too important not to bring to market. They’re too meaningful for our customers. We’ll just have to figure a way to allow this technology to have its own separate, unique track, or path into a vehicle. That’s what we started with as an objective in 1996. I wish I could tell you with certainty that we could have predicted that we would be successful, but it was really as much of a test on that dimension of the business as anything else we did. Seven generations of hardware later, Gen 8 is on the drawing board right now to go into production next year. We’ve been able to actually exceed some of the expectations we had for speed.
AC: OnStar has announced an upcoming collaboration with MapQuest, OnStar Web Destination Entry, to enable subscribers to plan their routes on a computer with MapQuest and then send that route data to OnStar. Can you go into the opportunities or advantages you see here?
CH: It will work for all Gen 7 systems as well, so there’s not a requirement for next-generation hardware. So all the vehicles we’re building right now, once we turn the system on live in production, it’ll be useful to anyone that has a Directions and Connections service plan as part of their OnStar relationship with Gen 7.
The way it works is, you will go to Mapquest and be able to determine a destination. Hundreds of millions of times per month, people are familiar with using the Internet to plan for their travel, and Mapquest is clearly the preeminent brand in travel planning, particularly for drivers. So you’ll use that kind of very familiar interface to search for a destination. It’ll plot that destination on a map, and then the other planning that you might do around that, like look for hotels in the area, look for scenic points of interest, or anything else that you might be interested in, you can clearly do that, because again people are very familiar with how to use that, then there will actually be a link on that page that will say “send to OnStar,” and you can click that link and send it to us.
Now you haven’t set the origin of your trip because, part of the beauty of OnStar is you get to start that trip from wherever you want, whenever you want. All you’ve really done is say where you’d like that trip to end. Literally, you’re sending that destination information to the OnStar secure server farm, linked to your account. As you’re driving, at whatever point you want to start that particular trip, you press the button in the vehicle. We would know that you’ve stored a destination, so you wouldn’t need to go to a live advisor for that particular Turn-by-Turn Navigation session. The voice would come on from the OnStar service infrastructure that would say, “I see you’ve got a web destination stored, do you want directions to that destination?” You’d say yes, it would then calculate, based upon your current vehicle location, the destination route and download it to the vehicle, and you’re off to the races.
AC: OnStar has voice recognition?
CH: It does. And as we always do in executing our services, we’ve tried to keep it simple to use. What we’re not trying to do with this voice-user interface, I don’t think the technology is quite ready for that yet, we’re not trying to do the complete destination determination with voice rec. What you’re really doing is just retrieving something that you’ve stored, you know it’s there. Our current thinking is, at least for the pilot that we announced, you can store up to five destinations and name them whatever you’d like to name them. Within the realm of voice-recognition technology, being able to accurately pull from one of up to five stored is a much more straightforward proposition than just purely ad hoc voice recognition.
AC: OnStar is sticking with audio commands, live or automated, versus maps displayed on screens as with some other automotive systems, aftermarket units, and portable navigation devices. What’s the reasoning here?
CH: If you understand how we’re trying to come at this business, our view has been for some time and it is now actually playing out — this is the year — an OnStar device will be built in every vehicle that General Motors sells in the U.S. and Canada. That’s literally an announcement that GM made two years ago, that we would put OnStar hardware and service in every retail vehicle sold in the U.S. and Canada of the 2008 model year. That’s playing out to a very great extent this fall, which is when the 2008 model year starts. What that says for us is the infrastructure we’ll have to leverage there, there’s not a screen in every vehicle, there’s not an embedded nav capability other than what we will build in this one-each-per-vehicle execution with OnStar. Part of the reason we’re not bringing a screen into every vehicle as part of OnStar is cost, but part of it is our research would tell us, the broadest preference for delivery of the occasional need for navigation, which is not to have a screen in the vehicle which someone looks at, but to be naturally taken by a well-executed voice prompt.
The other thing is, as at least GM is executing its screen-based nav systems today, which we build in a number of vehicles, we don’t allow destinations to be set in those nav systems while you’re driving. The utility of the on-the-go customer for navigation of a nav system, while it’s good for some people in some use cases, honestly, by virtue of being very careful about the driving experience, the whole notion of making sure that you architect whatever you do in a way that respects the fact that the person using the device is actually driving the vehicle — we’ve got a point of view that says the most responsible way we know we can bring this to bear for someone while they’re driving is hands-on-the-wheel, eyes-on-the-road execution, which is done with our turn-by-turn approach.
AC: Who makes the GPS receivers that will go into the 2008 model year vehicles GM will sell in the U.S. and Canada?
CH: We have two main tier-one suppliers of the hardware device that we call the OnStar unit. One is LG and one is Continental, which used to be a division of Motorola, which was recently spun out of Motorola and sold to an automotive supply company called Continental. [OnStar confirmed its GPS supplier is SiRF.]
AC: What opportunities or threats do you see for GPS in automobile navigation over the next 5 to 10 years?
CH: Our research would tell us that services like OnStar will increasingly be the price of admission for serious vehicle consideration, at least in North America. If today we’re building 25 percent of the share of the vehicle market in the U.S. and Canada, and we are the predominant supplier of these kinds of services today in this market, I would say there is significant growth potential for GPS solutions in vehicles as it would relate to the competitive market — basically, bringing to bear the pressures that we see coming, that will require folks to have services like ours.
I think there’s going to be a lot of growth, obviously. There’s a lot of discussion around can you use the GPS signal for some more automated type of vehicle control capabilities in the long-long-term, things that would help collision avoidance, and some of the things that might help with other safety kinds of opportunities. That’s probably a little bit more over the horizon based on some of the technical challenges that would have to be solved, as well as the existing installed vehicle population. Obviously it’s going to take some time for that to turn over, the kind of technology that you could rely on, every vehicle you might be driving near having that kind of capability you could interact with, so there are some issues that ultimately are going to need to play themselves out.
There are increasingly interesting possibilities to take things like traffic flow information, particularly in congested metro areas, to use the location of the vehicle, maybe to provide proactive traffic alerts to people. There, GPS might actually be the technology that allows for that data to be more accurately and broadly captured.
AC: Other outfits are exploring this. Do you have plans or talks underway to integrate these efforts into your enterprise?
CH: We’ve got a pretty good history of collaborating where it feels like it makes sense for us. We’re clearly not fixated on vertically integrating in every part of our business. But we also have a pretty clear idea of where we can add unique value and where our experience and insights will give us an opportunity to be competitively advantaged. We don’t have any specific announcements to make regarding traffic. But it’s very fair to say that traffic, as an example, has been an interesting service extension for us to consider. It’s not necessarily a broadly felt need across our subscriber base, but there are certainly significant subsegments of our subscriber base that would have interest in a traffic solution, if it was presented accurately, if it was easy to interact with, and if the price point of the ultimate solution made sense.
RSS
Become a Fan
Join our Network
Follow Us
Please send any technical comments or questions to our webmaster.