[sustran] Re: World Throughput Update (was Re: Jakarta busway twists
and turns?)
John Ernst
itdpasia at comcast.net
Fri Dec 14 06:29:24 JST 2007
Dear Todd,
At 11:10 AM 12/5/2007, Todd Edelman, Green Idea Factory wrote:
[...]
>> In a city like Jakarta, where for 20 years streets have been
>> generally filled curb-to-curb, it is very hard to accept that a lane
>> that looks empty could be carrying 10 or 20 times as many passengers
>> as the ones that look chock-full. You can imagine the pressure on
>> the police officer on the street.
>>
>HOW about a large portable L.E.D. sign - similar to ones warning of road
>hazards, temporary speed limit, etc. - which shows current passenger
>throughput? It could be based on average occupancy of private cars and
>buses... one such apparatus could be moved to different locations, would
>look great on TV, etc... there could be a live throughput counter on
>municipal website, local news websites...
Just to let you know, I like your idea for a variable message sign
showing passenger throughput. and have been thinking about how to implement it.
The critical aspect is that the sign at least appears to be giving
accurate information. (Jakarta has had signs for years allegedly
giving current pollution readouts, but they are either displaying
crazily optimistic numbers or simply not working at all. Neither is
effective.)
The ideal is a sign that would be giving exactly accurate
information, but that is not possible since there would need to be
some estimation of the number of passengers in each vehicle. This is
not a big problem as averages could be used for private vehicles, and
adjusted boarding counts for the BRT.
The counters could be an optical recognition system using video
cameras. Have not checked the prices, but am thinking Mitsubishi may
want to donate.
One problem is using optical recognition to count motorcycle traffic
of the bunched and weaving variety common in Asia. If anyone knows
relevant software or has other suggestions for this, I would be interested.
Thanks,
John
More information about the Sustran-discuss
mailing list