What's elliott?
About elliott
Contact us

t o p i c s

Business
Commentary
Destinations
Help
Leisure
Technology
Vault

s u b s c r i b e

Elliott's E-Mail, a free weekly newsletter, is your insider resource for moneysaving ideas.




• Read back issues. Like what you see? Now you can become an underwriter.

a l s o

Referring sites
Public relations
Visit Tripso
Home


s e a r c h

• Find a story.



Copyright Elliott Publishing. All rights reserved. For more information, call (305) 453-4781 or send e-mail to us.

Lost in Cyberspace
The Travel Critic · October 5, 1998

Getting lost is easy. Especially if you're navigating online.

Case in point: my recent road trip to Virginia. Driving directions from a travel site called MapQuest assured me the best route from my office in Annapolis, Md., to a business meeting in Arlington, Va., was through Washington, D.C.

Two more turns, and I found myself in a neighborhood with boarded up buildings and dubious looking characters lurking on the sidewalks. I couldn't tell where I was. When I finally got to an area with signs, they were in Chinese.

It took almost two hours to get to Arlington. It usually takes just over half an hour.

Hits to mapping sites have doubled in the last year. As the visitor traffic revs up, motorists are discovering the pros and cons of taking their driving cues from a computer. People who get lost like I did are finding that it's often their own blind faith in technology leading them astray.

"This is definitely an imperfect science," admits Marc Haverland, director of Internet engineering at GeoSystem Global Corp., which operates MapQuest. "The effort involved in verifying the accuracy of the data is immense. It's like shooting at a moving target-those maps change all the time."

Calling the mapping process an imperfect science is like referring to economy class as slightly cramped. Once you understand how a computer comes up with the turn-by-turn directions, chances are you'll think twice before trusting the Net as completely as I did.

As Haverland explains, maps aren't stored as maps, but as "segments"-each road represents a segment, which is tied to another segment and forms a whole map. When you type your starting point and destination, the computer calculates the fastest route based on these segments. It can't take into account geography, construction work, common sense or a lifetime of experience.

That's how you get creative-and at times misleading-directions. Like the ones that tell you to get off one highway and then get back on the same highway a few miles later. Or the ones that direct you to take a left on "unnamed road." And, of course, the directions that got me into trouble when I turned into a maze of sidestreets in the nation's capital.

"Although much of our road information is extremely accurate, some is not," reads a disclaimer on Maps On Us, another travel site. "The good news is that our routes can be very good. At times, we've found routes that are better than ones frequent travelers have used for years. The bad news is that sometimes our routes can be, well, creative. For example, a route might tell you to get off an interstate highway, take a local road and then get back on the same interstate."

It concludes by asking users to "be patient with any errors you may find. Our routing will continually improve as the quality of our data improves and as we invent better routing algorithms."

David Gersh, business development manager at Switchboard Inc., which runs Maps On Us, says the system is "not 100 percent foolproof right now. We're still working on the technology. We need to get a better understanding of how certain segments relate to other segments before it can get better."

OK, so the directions don't always work. So? They're free, after all. The trick is to know when you can-and can't-trust the computer.

Say you're flying into a really confusing city, like Boston, and you have to get from the airport to your hotel. Should you even bother with the Internet? Yes. But it's a very qualified yes, even for Gersh and Haverland. The computer is virtually dead-on accurate when plotting a course from point "A" to point "B" on major highways. It will also draw a map that can be printed out and used as a reference, according to both experts.

Everything else is a crapshoot. Exit ramps confuse the computer. So do county lines (because segments end and begin there, causing the system to reset itself).

Haverland is also concerned about the accuracy of the data of smaller streets. He remembers one road that showed up on MapQuest more than a year before it was actually built.

"I thought that was kind of funny," he says.

Christopher Elliott is a travel commentator and author of A Bridge to Nowhere: A Year in the Florida Keys. All e-mailed questions may be edited, condensed or republished at the site's discretion.