Most of the time, you can get along fine without in-car GPS. Your daily commute is marked by well-worn drudgery: You drive to work, to the store, and back home, rote trips for which you don't need help. And nowadays when you are lost, your phone can probably assist you. So it's no surprise that GPS firms are suffering. This week, shares of Garmin, the once-high-flying market leader, plummeted after the company lowered its revenue expectations for the year and delayed the launch of its long-promised smartphone, a device investors hoped would unshackle Garmin's fortunes from the apparently sinking GPS market.
But a few months ago, a Silicon Valley start-up called Dash Navigation put out a product that could well revive the sagging business. The Dash Express navigator packs a killer feature that other GPS systems lack: the Internet. Network connectivity powers Dash's primary attraction: what the company calls "crowd-sourced traffic." As you traverse your favored metropolis, the Dash Express anonymously transmits information about its location and speed to a central server. Every other Dash driver does the same. Using this data, Dash can paint a stunningly accurate picture of traffic patterns. Have you ever been stuck in a jam and wished there were some way to look two miles ahead to see whether things are still ugly? Dash essentially does that for you.
I've been testing the Dash Express for a week, and I'm floored. One morning rush hour this week, I drove from my home in San Francisco to Stanford University. At the start of the 30-mile trip, I plugged my destination into the Dash Express. The device gave me three possible routes, each with an estimated travel time based on traffic conditions gleaned from other drivers currently moving down those roads. I chose what Dash said was the fastest route, a straight shot down the congested 101 freeway. The device guessed I would arrive at Stanford in 59 minutes. Sixty-two minutes later, I was there. Along the way, the Dash predicted nearly every hurdle along my trip with eerie accuracy: Traffic slowed down just where the color-coded map showed yellow, orange, and red roads, and speeds picked up again exactly where Dash's map was painted green.
The Dash isn't perfect—its navigator-lady voice-over has terrible pronunciation skills, and its software and hardware isn't nearly as sleek as those in Garmin's GPSes. Plus, at $299 for the device and at least $10 a month for traffic service (after three free months), the Dash Express isn't cheap. Yet its ability to predict traffic and, most important, guide you around congestion makes it a must-have for commuters. Dash transforms the GPS navigator from something most people rarely need into one of those revolutionary applications—think TiVo, the iPod, Napster—that you can't imagine doing without once you've tried it.
It's true that traditional GPS devices can be souped up with traffic-alert services. (Most receive traffic data through an FM-radio receiver; Dash connects to the Internet through open WiFi and cellular connections.) You can also get a picture of road conditions by consulting online maps; the iPhone's Google Maps application colors many highways in red, yellow, or green, for instance.
But the Dash's system is much, much better. First, it's more accurate. To predict conditions on your route, traditional GPS devices and Web maps rely mainly on what traffic scholars call "incident data." These systems get updates about car crashes, road construction, and other slowdowns and then estimate how fast traffic might flow around the holdups. Some systems add information from sensors implanted in major roads. But sensors are also imprecise. In traffic jams, cars move in a stop-and-go pattern—and if the sensor happens to be located just under the "go" portion of the jam, your GPS device will think the road is much kinder than it really is.
Dash also receives incident and sensor data, but it adjusts all its numbers with on-the-ground conditions fed back by real drivers. The system uses this info both to plan your route and to suggest changes as you're driving. If Dash senses a sudden slow-down ahead, it will ask whether you'd like to be routed around it. Sometimes, it will even guide you off the freeway and through surface streets, for which Dash also knows traffic conditions. (The system tracks traffic patterns over time, compiling a database of how quickly all roads move during 672 discrete intervals during the day.)
There's an obvious chicken-and-egg problem with Dash's system. In order to get good traffic data, Dash needs a lot of drivers—but to get a lot of drivers, it needs good traffic data. The company argues, though, that because many drivers follow similar routes, it can achieve a critical mass relatively quickly. In an average-sized metro area, Dash needs only a few hundred drivers before most of its data is coming from the crowd, says Mark Williamson, Dash's director of product marketing. In the largest areas—New York City and Los Angeles—Dash needs only a few thousand devices to get a good picture of traffic. (Dash won't say how many drivers it has in each of those areas.)
Dash's Internet connectivity helps with things besides traffic. Traditional GPS devices ship with databases of millions of shops and attractions across the country. Like a printed phone book, these databases go out of date: If you bought your GPS a couple of months ago, for instance, it will think there are 600 more Starbucks in the country than there now are. Over time, as roads shut down and new developments spring up, maps go stale. In order to refresh your device, you've got to buy an update disk.
Dash updates itself automatically with the latest maps, and it offers something an order of magnitude more useful than a built-in database of attractions: a Web-based search engine. When you look for nearby shops in Dash, you're really searching Yahoo, which already knows about all those shuttered Starbucks.
For all this great functionality, Dash faces a major vulnerability as a business proposition: Many of its features can be replicated on smartphones. Technically, the iPhone can do everything Dash does—it's got the Internet, GPS, and a touch-screen interface. It's possible to imagine another start-up building a Dash clone on Apple's device or on any other advanced phone. Considering how many of them are out there, the crowd-sourced traffic information generated by the iPhone would put Dash's data to shame.
Williamson told me that the company is keenly aware of that possibility. For now, he says, Dash is offering its service on only its own GPS device, but he did mention the possibility of porting it to other gadgets, like the iPhone.
In the meantime, the traffic data that Dash learns from its drivers could also prove valuable. The licensing possibilities look lucrative—Google, Microsoft, and Apple might all want better traffic data for their maps products. UPS, FedEx, and the Postal Service could probably also do with a clearer picture of road conditions. And Dash might even be able to help Starbucks out. At a recent tech conference, a Dash executive pointed out that Dash knows where people drive and knows where people search for coffee. That means it knows exactly where Starbucks should open up its next location in Arkansas: Highway 40, between Little Rock and Memphis, Tenn.