@phdthesis{oai:ynu.repo.nii.ac.jp:00007692,
 author = {Saadia, Tabassum},
 month = {Sep},
 note = {Now-a-days many developing countries are suffering from high motorization trend that has produced several unwanted by products. Though private vehicles provide high level of access and freedom but have enormous associated risks in the form traffic congestion and accidents, noise and air pollution and high energy consumption. Admitting these facts, many developing countries have initiated efficient mass transit systems to relieve these side effects. These systems provide a comfortable, safer and efficient services and so can be fairly preferred over private vehicles. Also the target of these systems is same as to cut down the private vehicle usage by attracting commuters from diverse social income classes other than low ones. However most of these systems are planned and designed independently, leading to the difficulties in accessing and egressing. Accessibility of these systems, which is one of the basic components of performance measure used for transit (Wardman 2004), is still crucial in most of the developing countries. Lack of any regular feeder services creates gap between ones’ origin and mass transits station that cause dissatisfaction to commuters and they lose patronage towards these systems. Currently paratransit is emerging as dominating feeder mode in several developing countries and connecting commuters from residence to public transport systems. However these paratransit have certain service areas and potential passengers. Also most of these are informal, not well organized and unsatisfactory and their movement is restricted to certain parts of the cities. Therefore a detailed evaluation is essential, also from commuters’ viewpoint, before implementing them as feeder. Unlike paratransit, a regular feeder bus, a form of public transport, can provide safer, comfortable and reliable service and will be free from the enormous risks as associated with paratransit. However a regular feeder bus requires a careful and vigilant planning and design. The precise selection of potential demand locations and appropriate approach to generate optimized feeder routes can result in the development of successful feeder bus operation. Optimized feeder routes can reduce the overall journey time of commuters while travelling in a comfortable mode. The purpose of the feeder route is to connect all the destinations for which demand generates from the mass transit station. All the previously proposed models for feeder routes generation are applied through conventional engineering approach; in which routes are generated separately for each station by assigning certain route generation nodes to a station. However a route generation node is a part of network and might be connected by some other stations too, its distance or travel time might be improved by inserting it in any other route for any other station. This issue remains unknown in typical approaches and some improvements are required to consider all or multiple consecutive stations simultaneously. Therefore this research intends to improve the conventional application method of route generation which can help in decision making by providing a set of solutions. The improvement in the current approach also aims to reduce the overall route network length to lower the operators’ cost. For this research, Lahore; the second largest city of Pakistan is selected as the study area. Lahore has a mass transit system in the form of Bus Rapid Transit System (BRT). Currently no regular feeder service is provided with the system and only existing modes are being used for feeder purpose. The existing modes are comprised of conventional public transport (public buses) and informal paratransit (auto rickshaw and motorcycle rickshaw). The role of auto rickshaws as feeder is almost equal to none being costly and fewer in number, whereas motorcycle rickshaws have made deeper inroads and are being used as feeder most frequently. In order to understand the current travel pattern and characteristics, two kinds of field interview surveys were conducted during Feb-March, 2015, with the help of university under graduate students. The first type of survey was conducted at few BRT stations comprising of 311 samples while the other was conducted nearby offices, workplaces and activity centres which made up 296 samples. The collected questionnaire data was analysed step by step and finally structural equation modelling techniques were applied. The results of the study revealed that currently major BRT users belong to low income class and those who own no vehicle. Among these users, work commuters are significant of all, which is also due to the location of BRT corridor that provides direct access to many workplaces. The commuters who own private vehicle and belong to other than low income class are more prone to use their own vehicle than BRT, even if they can make their egress trips by walk only. As far as the feeder mode is concerned, majority of BRT users are those who can make their access or/and egress trip by walk only. However the share of walk for the egress trip is dominant due to the proximity of workplace locations from BRT stations. After walk paratransit has greater share than public transport. For public transport, results of spatial analysis reveal that majority of city’s population is living far from the nearest bus stop and so cannot avail public bus as feeder. Whereas for paratransit commuters perception show that commuters are most dissatisfied by the vehicle quality as it offers bumpy, uneven and risky travel and also the service reliability. Further results show that in order to improve the BRT accessibility, all these dissatisfied factors must be dealt. Replacement of paratransit with feeder bus is a desirable option, also commuters show strong willingness to pay for the regular feeder bus. The study further suggests few preferred towns to initiate feeder service based on residential landuse type, owing higher density and uniformly populated. It also introduced an improved approach for optimized feeder routes generation. In the proposed improved approach feeder routes are generated, simultaneously for multiple adjacent (or for all) stations. The proposed approach yielded better optimization with lesser computational time and effort. The altogether insertion of all demand points for all the selected stations not only minimizes the average route length but total sum of all route lengths as well. This mechanism is found to be beneficial in achieving better trade-off between users and operators interests, when compared with the conventional approach. Therefore it is recommended to apply proposed approach to gain better optimization results. Based on the study findings, other developing countries must consider local differences and service area and passengers for various types of paratransit. In the end recommendations have been made for future work by proposing constrained multi objective optimization and incorporating other social income classes for evaluating paratransit.},
 school = {横浜国立大学},
 title = {Feeder network design to access an existing bus rapid transit system in Lahore},
 year = {2016}
}