This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
The topic of urban bus scheduling has been receiving popular attention since 1960s. Growing urban population and the new developed residential areas constantly pose a challenge to cities like Singapore, as Singapore steps up its effort to cater for the transportation needs of these new people. Singapore is experiencing an annual 1% population growth, and there are evidence made us believe that its public transportation sector is reaching the bottleneck capacity. Singapore transportation authority is at an urgent move to improve the capacity of the buses and trains and to provide a safe, convenient and quick service to the commuters. Starting with all the current bus planning strategies in the literature, this project aims to look for both theoretical and simulation guidelines for policy-setting, as well as to suggest specific configurations that tackles a particular demand pattern.
Regarding literature in urban bus planning, the most basic approach is to set up standard spacing for transit stops. The current guidelines given by Transportation and Research Board of United States are 500 to 1200 feet for urban areas and 300 to 1000 feet for Central Core Areas of CBD (Transportation Research Board). However, different research differs on the exact optimal bus stop distance. Vaughn and Cousins have shown that in the many to many demand case the spacing of minimum travel time is inversely proportional to the square root of the demand for boarding and alighting. Levinson(1983) concludes that performance of the bus system can be improved best by keeping the number of stops to a minimum.
As in headway control, there are few analytical studies compared to the simulation studies. It is known and proven by Osuna and Newell (1972), that the even without too much traffic disturbances, high frequency transit will eventually destabilize the system and make the bus bunch. The rule of thumb is to add time slack for the bus arriving timings. However, as pointed out by Newell (1977), even adding time slack is not able to ensure that the performance is steady and stable. Thus there are two ways proposed to eliminate bus crunching sudden disruption: adding multiple control points by Daganzo(1997); and dynamic holding times based on real-time headway information by Daganzo(2009).
Moreover, with the readily available real time bus arrival forecast technology such as Iris NextBus Service, the forecast may change the customer arriving pattern from exponential arrival time to constant arrival time. And with the help of the simulation tool, this study aims to find out the effect of real time forecast tool on bus system performance.
Also, it is shown that under bus capacity constraint, bus operating company prefers to have a large number of bus stops( narrow bus stops spacing) policy, with the bus stops spacing far less than the optimum solution obtained in part 1), which has no bus capacity constraint. However, the down side associated with increasing number of bus stops is the decrease in average number of people that is carried per time unit. In other words, in this context increasing number of bus stops does all good except it decrease the actual carrying capacity in terms of number of people per time unit.
However, it is shown that settings which have narrower bus stops spacing (greater number of bus stops) needs less number of buses to achieve the service criterion. I think this conclusion could be flawed, because the conclusion may be confounded by the reason that passengers in more bus stops setting have a smaller average trip length compared to those in setting with fewer bus stops, not because of it has more bus stops. In other words, due to the incompleteness of the model, the conclusion reached here may be caused by another confounding factor, which is the average trip length, but not the factor itself (by which I mean â€œhaving more bus stopsâ€Â). To illustrate, the average trip length for 16km line, 10 stops is 8.8km; whereas for 5 stops it is 9.6km. Since the setting with 10 stops have a shorter trip length, the passengers in that case will travel faster on average and the occupancy rate on the bus are lower than that of a 5 stops case. Thus fewer buses are needed to achieve the same service criteria. Therefore, it can be concluded that average trip length is indeed a possible cofounding factor.
As people tend to depart from some stops (i.e. residential areas) and alight at some stops (i.e. CBD), some travel pattern are formed. Because of the travel patterns, there exists the imbalance of arriving density, namely the arrival rate at some of the bus stops are substantially larger than that of the rest. Moreover, this causes clustering of the buses, which lead to extreme long waiting time and waiting time variations. (However, the timeliness of the pattern is not considered, in which I mean the model does not consider whether it is flow this way or that) To tackle this problem, uneven bus headway by varying bus stop density or express buses has to be introduced.