Early praise for Google Maps' bike routes

Improving Software series banner

The nice thing about the Internet, or so I've been told, is that it has all this information. Perhaps you've noticed this lately, but the big problem has been that there's no one way to get at this information with any kind of consistency.

Supposedly Google is the "portal" for most of the world's information, which may be why so many people find Betanews by typing "Betanews" in Google. In one respect, you might expect Google to have an interest in creating that consistent methodology for getting at information. On the other hand, given that so many folks depend on Google Search just as it is now, you could see how Google might very easily come to the conclusion that there's no new benefits to be gained through improving its software, just to keep the user base it already has.

My friend and colleague Carmi Levy touched on that point this morning in Wide Angle Zoom, where he turned a reluctant thumb down on the Google Buzz social network experiment. You can call it "beta" or, as Google strangely decided this time around, not call it beta, but after using it for a while, it's difficult not to come away with the opinion that this is indeed an experiment...and you're the subject.
But as Carmi also noted, different Google services are so inconsistent with one another that they may as well have been developed on different planets -- which suggests that Google may have been copying the Microsoft development model after all. Perhaps the best case-in-point example comes from Google Maps. Yesterday, to the delight of many, it added bicycle routes to its plotting options.

If you're a biker, you know that there are peculiar distinctions between the routes you plan as a pedestrian, and those you use on a bicycle. In cities whose planning commissions have added lanes to boulevards by destroying sidewalks, you know that unless you enjoy motocross and frequent encounters with beat cops, you can't plant bicycle tires on the same places you can plant your feet. I grew up decades ago in Oklahoma, at a time where for most areas of the state, "downtown" was denoted by the presence of the stoplight. Before I could drive a car, I biked four miles back and forth from my first high school, and (sometimes) the ten miles back and forth from my second, at a time when that was considered a normal way for a kid to get around town. Even then, I distinctly remember there were spans of a few hundred feet or so every so often where, if the paved road ended and the granite rock paths began, I'd pick up my bike and tread on foot between people's houses. It helped to get acquainted with them first (see: "beat cops").

The brilliance of the Google Maps experiment is that it accumulates the data that Google has gathered, not just through map scanning and satellite imagery and those strange cars you see cruising every little granite path, but through its advertising service as well. As a result, from day one, the bicycle route part of its service is capable of guiding cyclists using the tools that matter to cyclists: the landmarks they see along the way -- the gas station, the church with the tall steeple, the Italian restaurant. No service established exclusively for the purpose of mapping the world, for motorists or cyclists, would have ever assimilated information at this level of granularity; only through co-opting the advertising and location database with the mapping database could this ever be feasible from a business standpoint.

When I tested Google Maps last year as a pedestrian/public transit dependent in Los Angeles, I discovered it was using a form of "reverse tunnel logic" to compile its suggested routes. Quite literally, it would have me traveling by foot one mile south to catch a bus that would take me 1.1 miles north; and it also would have had me scaling viaducts where there were drainage ditches below and barbed wire above. I pointed out to Google at the time that Maps appeared to fail to consider time and effort consumed as factors in assimilating its routes. The response I got at the time from Google was, thank you for your input but, well, you can't know everything about everything.
As I've learned from experience, yes, you can. Living in Indianapolis, for example, I know there is a kind of "superhighway" exclusively for non-motorists, formed from the brilliant idea of paving over a disused railroad track. Named for the railroad that used to own it, it's called the Monon Trail -- a 16.7-mile stretch of track that's well kept, fairly well patrolled, and is the most foot-friendly stretch of asphalt I've ever traveled.

Since it bisects most of the city north-to-south, there are a multitude of ways someone could get to the Monon Trail. Indianapolis is notable for having plenty of bike trails in certain areas of the city, and no way in hell a bike could get through in others. For my initial test of Google Maps for cyclists, I wanted to see whether it would make the same decisions I would, having lived here now for 18 years.

With pedestrian maps, Google's suggested route is plotted in blue, with the relatively foot-friendly paths outlined in solid green, or dotted green for "better than most." If I wanted to walk from my house all the way to, say, Conseco Fieldhouse downtown where the Pacers play, Google Maps would try to plot the most direct foot route it could. Even though I know for a fact that walking the Monon Trail is much, much faster than down the side of a boulevard, the fact that the Trail is a half-mile out of my way means that Google's walking map won't suggest it.

It is a very different story for cyclists, and here it's clear that someone at Google listened to my advice. It takes a little more effort to pedal west a half-mile to get to the Trail, but once you're there, the effort pays off with a four-and-a-half mile stretch of paved roads, overpasses, and relatively safe crosswalks ("zebra crossings"). Tunnel logic evidently was not used to compile this route, but rather an assessment of the time and even luxury benefits to be gained from going out of one's way -- the type of assessments Google Maps did not make last year in L.A.

Here's what I mean: This screenshot shows the suggested route from one of my favorite neighborhood pizza joints to Conseco downtown. Now, Google knows that Allisonville Road recently added a bike lane (not a good one, it's in dotted green), that it leads to a nicely protected sidewalk down Fall Creek Parkway, and that would be the most direct route downtown. Indeed, if I were a pedestrian, that's the route it would suggest. But as a cyclist, I know that's not the route I want -- crossing onto Fall Creek over Binford Blvd. means running across ten lanes of traffic without a crosswalk, where motorists are speeding through on a shortcut to I-69.

So Google Maps takes my bicycle (a vintage Takara racer, if you're interested) a mile out of my way, through a residential area, just to get me to the Monon Trail. And that's exactly right -- there's no question it's a lovelier, easier, and in good weather, faster route.

Next: Seeing where you're going, and where you shouldn't go...

7 Responses to Early praise for Google Maps' bike routes

© 1998-2024 BetaNews, Inc. All Rights Reserved. Privacy Policy - Cookie Policy.