[sustran] Re: Fw: Fw: Re: Traffic Congestion in Bangalore (WITHATTACHMENT)

Chris Bradshaw c_bradshaw at rogers.com
Sat Mar 22 05:33:56 JST 2008


Counting trips on foot is indeed a challenge, one that most traffic 
engineers don't worry about resulting in undercounts.

I am aware of four different types.

1. The origin-destination survey, used to feed data to computer models, uses 
telephone surveys to ask people what mode is used for different trips.  The 
first problem is the fallibility of memory, and the second a certain 
political correctness that can cause respondants to be less than truthful. 
And thirdly, there is the sampling used in choosing a sample, or the trips 
they query about, or the date for the survey, etc.  And the coding, as was 
mentioned, doesn't allow for multi-mode trips to be recorded as such.  For 
instance, I estimate that transit trips are made up of 50% of the time 
duration to be in walking and waiting, for buses and don't-walk signals.

2. The second is cordon counts, in which the mix of travelers is recorded 
across a particular line on a map.  These lines are chosen for simplicity, 
where the traffic patterns are least complex, e.g., green corridors, rail 
lines, rivers/bridges, etc.  Not only do people on foot avoid trips across 
such empty spaces, but walking trips across these are not going to be long 
enough to cross a second cordon in the same time vs trips by car.  Walking 
is undercounted, in part, because the trips are shorter.

3. The third is the intersection count.  The vehicular modes are all counted 
simply, but pedestrians are problematic.  Those turning right, and therefore 
not leaving the sidewalk, are not counted at all.  But, those who have to 
use two intersecting crosswalk to complete their trip might be counted 
twice.

Of course, trips using only local streets, by any mode, are rarely counted 
at all, partly because these streets are rarely used that much.  Only when 
neighbours complaint about "cut-through traffic" do cities count them, and 
then they ignore walking, although the presence of pedestrians is relevant, 
but in the opposite way: pedestrians represent both the general positive 
presence of "eyes on the street,": (a la Jane Jacobs) and become part of the 
'traffic calming'; but also a measurement of the human endangerment that 
such traffic poses, not to mention a measurement of how bad the car problem 
is, since the more pedestrians, the less serious the perceived threat.

4. Finally, there is the matter of "warrants" for pedestrian crossings. 
Here traffic engineers count the number of pedestrians only that _attempt_ 
to cross a roadway at a particular section.  The higher the count, the 
greater the warrant (the justification) for providing a special combination 
of paint lines and perhaps a signal or extra lighting to help pedestrians 
cross. The problem with it, is that where such crossing assistance is most 
needed is probably where few crossings are attempted and where pedestrians 
long ago found other routes or, equally likely, other modes of travel. 
Where there are plenty of people crossing is where little improvement is 
needed, although often there can be problems with pedestians with special 
needs.

Chris Bradshaw
Ottawa




More information about the Sustran-discuss mailing list