pardon the long post. this turned out longer than imagined.
naveen that is a comprehensive list. a good reference list. i think i know where you are headed. some months ago i started a small list on metro. abandoned it midway after a system crash. i wanted to compile population, total network length, network length by mode, ridership by mode for a few cities. a few heavy cities, few medium ones and a few light ones. then we can see if the network density has any effect on effectiveness of public transport.
i have a suspicion that irrespective of the mode, network length is important for effectiveness. not just in the obvious way, but in terms of self sustenance and effectiveness. just 1 line here and another there even if on the densest corridor will not cut it. ridership is a function of network size. coverage is very very important.
in that list i suspect some cities with just HOV lanes to show have sneaked in. if we do effectiveness trips_by_buses/total_trips type of tests, i bet half these cities will come cropper. for example, in the us there are quite a number of cities which are trying to solve the rush hour problem and have self righteously come up with these HOV lanes without doing anything about transit itself. things like car pooling are outcomes of that type of thinking. cute sounding, but nothing more. how many cars do you have to pool to solve hosur road problem?
bus systems in the us atleast are a big joke. if you follow the metrics and usage stats in fha reports on bus systems and look at their metrics, you will die laffing. laskmeshwara in gadag district has more ridership than what these folks are trying to achieve. ok rhetoric, but the point is, even if i may not have been comprehensive, i have not seen any designs in fha reports capable of handling more than 20 people at a stop. and these guys go by the book. in the pic SB posted there are more than 20 people in the middle of a bridge waiting for a bus.
but bus systems have their advantages. infinitely flexible and reconfigurable. thus can be very responsive. and they have a reach which ROW systems can only dream of. but they are not for long hauls and for high volumes. It is a misnomer to call what ever they are going to do on ORR as BRTS. RITES' prescription is to reuse existing intersections rather than redesign them to facilitate BRTS, when the entire game of BRTS is in intersection design to achieve quazi ROW. midblock ROW is fine, but if you cant negotiate an intersection in good time, there is no way this thing is going to be rapid. and unless it is rapid, there is no way it can serve as a good long haul solution. Average trip distance in bangalore is 9kms. for reference, silkboard to electronic city is 9 kms. at the far end of hosur road, at points south of electronics city, there are 30k vehicles. how many buses do you need to solve hosur road? what about the mother of all high volume roads in bangalore, bellary road? yeshwantpura or even ORR?
again, i may be wrong. but murali and vasanth (god bless their souls) have been dropping crucial data points on buses in blr -- whether it is bmtc abandoning grids, or ellideera. these might be outliers, but i have been making a living out of saying that outliers are nothing but un-modeled dynamics so i have to make my case on this one. here is my HYPOTHESIS.
bmtc is finding it difficult to do long haul. with increasing congestion it has become very difficult for it to perform at any meaningful level of service. grids are taking long and ellideera is revealing embarassing stats, for example. so they dropped both of them. much of all this is due to the reality of blr. there is a very high volume of private vehicles. as many imaginary pitched battles i have fought with erstwhile justice saldhana, i agree with him when he says that the palike cutting thousands of trees has not translated to even a 1% improvement for bangalore. on a similar train of thought, it would be very very interesting to figure out how many buses are there too. piecewise, adhoc solutions are not the way to go.
my basic diagnosis is this: we all know what the design levels of bangalore's roads. what this means is, per kilometer the entropy(chaos) levels on the roads are the same through out the city on an average. as you increase the trip length, entropy(chaos) adds up. which means any given metric, like reliability, trip times etc etc all have high variance, aka uncertainty. and high variance is bad news for metrics. i would love to sink my teeth on bmtc records, ticket sales, trip lengths, trip times etc etc and test this hypothesis against that. i am fairly certain, that my hypothesis will hold up.
and yet bmtc has concentrated on long haul. when the contributors of entropy is short haul trip count. here is a pointed question- would blr be better if bmtc concentrates of intra hub transit and let all inter hub transit to ROW systems?
