A New Performance Index for Evaluating Transit Quality of Service

This article proposes a new performance index called Transit Service Indicator (TSI), which could be used as a comprehensive measure for quantifying the quality of service of a transit system. TSI integrates multiple performance measures (e.g., service frequency, hours of service, route coverage, and travel time components) within a systematic framework. It takes into account spatial and temporal variations in travel demand, recognizing that quality of service is a result of interaction between supply and demand. A case study is conducted to examine the sensitivity of the proposed TSI to various system design and condition variables and parameters. Introduction The latest Transit Capacity and Quality of Service Manual (TCQSM; Kittelson & Associates 2003 ) is an excellent supplement to the widely accepted Highway Capacity Manual (HCM; Transportation Research Board 2000), with a systematic framework for addressing various conceptual and methodological issues related to transit capacity analysis and quality-of-service evaluation. One of the main features of the TCQSM is its adoption of the level-of-service (LOS) concept introduced in the HCM for measuring the quality of transit service from users’ perspectives. Journal of Public Transportation, Vol. 10, No. 3, 2007 48 As acknowledged in the TCQSM, quantifying the quality of service of a transit system is much more complicated than evaluating a highway facility because of the involvement of multiple players (e.g., transit operators, passengers, vehicles) and a wide range of interrelated factors (e.g., spatial and temporal coverage, comfort level, reliability). As a result, the current TCQSM has opted to use multiple LOS measures, instead of one or two measures as the HCM, to evaluate the quality of service of a transit system or its specific components. For example, six LOS measures are proposed for evaluating the quality of service of a fixed-route transit system, encompassing both service availability (service frequency, service span, service coverage) and service quality (passenger loading, service reliability, transitauto travel time difference). These measures reflect different aspects of a transit service system as perceived by typical transit users. One of the major disadvantages of using multiple LOS measures is its difficulty to provide an overall quality-of-service evaluation required for comparing different transit routes, travel corridors, or transit systems. The main objective of this research is to explore the possibility of combining some of the LOS measures into a single quality-of-service measure. The attempt to develop a comprehensive quality-of-service index is not new. Several past studies have made considerable progress on developing service indices to measure transit quality of service, as summarized in Table 1. Rood (1997) proposed a service availability measure called Local Index of Transit Availability (LITA), which includes three components: frequency, capacity, and route coverage. Hillman (1997) developed the Public Transportation Accessibility Level (PTAL) index to measure the access availability to the public transit network. Florida DOT introduced a new quality-of-service measure called Transit Level of Service indicator (TLOS), which is defined as the percentage of time that an average person can use the transit service (Kittelson & Associates and URS, Inc. 2001). This indicator incorporates the coverage, frequency, duration of the transit service, availability and quality of pedestrian paths to transit stops, as well as the number of people and jobs receiving transit service. All these indices model only the availability aspect of a transit system with no consideration on service convenience (e.g., travel time) and demand distribution. Polzin et al. (2002) were the first to suggest the need to consider demand distribution in evaluating transit quality of service. They proposed the Transit Service Accessibility Index (TSAI), which measures how well travel demand is served using time-of-day travel demand distribution to determine the relative value of the A New Performance Index for Evaluating Transit Quality of Service 49 Re fle ct in g Tr an sit Re fle ct in g Tr av el In co rp or at ed P er fo rm an ce Re fle ct in g Tr an sit Co m fo rt a nd D em an d In di ce s St ud ie s M ea su re (s ) Av ai la bi lit y? Co nv en ie nc e? D ist rib ut io n? Tr an sit L ev el o f Ki tt el so n & Se rv ic e co ve ra ge , f re qu en cy , s er vi ce Ye s N o N o Se rv ic e In di ca to r A ss oc ia te s a nd sp an , p op ul at io n, jo bs U RS , I nc . 2 00 1 Tr an sit S er vi ce Po lz in e t a l. 20 02 Se rv ic e co ve ra ge , s er vi ce sp an , Ye s N o To ta l # o f t rip s A cc es sib ili ty In de x fr eq ue nc y, tr av el d em an d Lo ca l I nd ex o f Ro od 1 99 7 Fr eq ue nc y, ca pa ci ty , r ou te c ov er ag e Ye s N o N o Tr an sit A va ila bi lit y Pu bl ic T ra ns po rt at io n H ill m an 1 99 7 Se rv ic e fr eq ue nc y, se rv ic e co ve ra ge Ye s N o N o A cc es sib ili ty L ev el Se rv ic e Q ua lit y In de x H en sh er e t a l. 13 v ar ia bl es (i .e ., t ra ve l t im e, Ye s Ye s N o 20 04 fr eq ue nc y, et c. ) Tr an sit T ra ve l T im e D ow lin g an d Tr an sit tr av el ti m e Ye s Ye s N o