Comments on: Where Is Michael Jones?

By: Peter Smith

Peter Smith — Mon, 16 Jun 2008 21:09:04 +0000

Thanks Jack,

You and Michael are problably right – I think I even misinterpreted what Michael was saying about paths.

The ‘type methodology’ idea is good.

Thanks for checking in!

By: Jack

Jack — Mon, 16 Jun 2008 19:51:00 +0000

I do agree with Michael T. that path directions are essential. At first paths can be treated in simple fashion, as simply another road except cars can’t use it (or in the case of sidepaths, cars can). More detailed directions like “stay left and cross the bridge” can wait until round 2 or 20, but even that’s not too tough if you make provision to tie a simple text field to each intersection or info point — the content can all be “crowdsourced” as you say). But even Round 1 should not be released without displaying major paths. In my suburban D.C. area, the hardiest road commuters still depend on paths ranging from long standalone rail trails to short connectors linking dead-end streets. I suggest a type methodology that describes basic type (road, sidepath, both, or standalone trail). Then you can worry about specific road accommodations (wide outside lane, bike lane, etc.).

By: Peter Smith

Peter Smith — Wed, 28 May 2008 11:33:34 +0000

Thanks for the comments Michael,

Though paths and those ‘move to the left’-type instructions are important, I would argue they are not necessary for ‘Round 1′ of bicycle directions functionality. There are many reasons for this, but perhaps the most compelling is that few of us have regular access to special bike/walk paths anyway. In and around downtown Austin, for instance, I regularly do not see/use special paths of any kind – I just try to stick to Class I and Class II roads – high and medium ease-of-use roadways, respectively.

And while not all of the data we could ever want is readily available, some of it is. The ‘Austin Bicycle Route Data’ link in the right column has road/class data for Austin (this ‘class’ data from city ‘bike maps’ already accounts for traffic levels, the presence of on-street bike lane markings, etc.).

ftp://ftp.ci.austin.tx.us/GIS-Data/Regional/coa_gis.html

And I would say there are several thousand cyclists in the U.S. and around the world who are more than ready to help crowdsource any required data.

The positive repercussions of a ‘Bike There’ option far outweigh the direct, immediate benefits to existing and soon-to-be cyclists.

Good luck at the Conference!

By: Michael T. Jones

Michael T. Jones — Wed, 28 May 2008 10:58:59 +0000

Hi Peter,

Cycling and walking directions are important, useful, and as you observe, beneficial to the health of those who follow them. What you don’t mention is that all navigation and routing services are built atop detailed data describing potential pathways.

For driving directions, this includes road networks, speeds of travel, directions of travel, lane restrictions such as commuter lanes and times, one way streets, time based turn restrictions, and many other subtle details. For public transit, such data is in the form of transit schedules.

The challenge in extending services to cycling and walking is in extending the coverage of the possible paths to be sufficiently detailed to capture knowledge about facts like “move to the left here and walk your bike across the pedestrian walkway” or “cut through the lobby of the Aoyama Twin Towers post office and go down the stairway to the subway entrance.” The related challenge is getting this information not just for a whole city, but for a whole planet of cities.

Many Googlers walk and pedal. As this data becomes available, yo ushould expect that we will be eager to make the most of it.

(Posted at 3:57am PDT, just before the Google I/O developer conference)