Problem generating valid Walk time polygons

JenniferMinnick

Hi -

I've spent some time with a great ESRI customer service rep on this topic but so far we have not found a solution so I thought I'd try here as well...

I'm using ArcGIS 10.1 (lowest level - whatever they are calling it these days) and the NAVTEQ Q411 streets data that comes with BA 10.1.

So, with the prior version of the streets network (the *.rs) file in order to get a walk time polygon we'd change all of the street speeds to 3 mph and we'd get a moderately valid walk time polygon.

The new version of the streets network and the new Network Analyst interface which utilizes (so far yet mostly undocumented) historic traffic speeds (for driving), and some vague sort of average for non-time dependent driving analyses.  There is an option under Layer Properties>Analysis Settings to set a "restriction" to "Walking"

The result of this shows that I can walk approximately 3-4 MILES in 10 MINUTES.  Nope.  No go there.

So the question is:

1) What am I doing wrong?
2) How is NA evaluating walking?

Please help me!!  A lot of things hinge on my obtaining correct walk times OD and "trade area" polygons.

Thanks,
Nif

[ATTACH=CONFIG]27551[/ATTACH]

10min_WALKtime_poly.jpg

Melinda Morang

"Walking" assumes you're walking.  In the background, I believe it's simply restricting streets based on an attribute in the street feature class that indicates that pedestrians are forbidden.  I agree that the naming convention is a bit confusing.  For a pedestrian analysis, "Walking" should be checked.

To eliminate one-ways, just turn the one-way restriction off (assuming there is one).  If one-way is unchecked, one-way values will be ignored.

If you solve your analysis using a distance measure (such as Miles or Meters), the solution will be optimized based on network distance rather than travel time.  The travel time impedance attributes are generally designed for cars, and they use the traffic data and speed limits.  However, the distance-based impedance attributes are ignoring all that and only using distance in the optimization, so it can be used for a pedestrian analysis.

The main limitation you will run into with your network will be pedestrian paths or cut-throughs that aren't included.  Pedestrians can walk on paths, across parks, through buildings, etc.  However, the basic street network with a distance impedance attribute will give you something reasonable to work with.  You could certainly make your own network dataset designed for pedestrian analysis if you have the data.