Cover Image

Not for Sale

View/Hide Left Panel
Click for next page ( 41

The National Academies of Sciences, Engineering, and Medicine
500 Fifth St. N.W. | Washington, D.C. 20001

Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement

Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 40
40 for such analyses may be in danger of sabotage. Agencies may 4.10.2 Transfers and Linked Trips not want to use the AVL data directly to discipline operators, but it can certainly be used to help dispatchers and supervi- While APCs provide the data needed to analyze demand on sors better target their efforts at conventional discipline. For a route, they do not capture the information needed to iden- example, some agencies report that if data indicates a recur- tify linked trips or transfers. However, if fare media include ring problem with a particular trip starting late, a supervisor unique IDs (as is the case with both magnetic cards and smart might be requested to observe. cards), stop- and time-stamped farebox transactions permit More seriously, safety can be compromised if operators analysis of transfers and linked trips. Linked-trip analysis is are punished for getting behind schedule. Such concerns do especially important in Canada, where linked trips are the not necessarily mean that operators should not be given standard measure of transit use. feedback on their performance. Experience with data col- Despite fare systems not capturing alightings by ID code, lected automatically in Rotterdam's tram operation shows there have been successful efforts in New York, Chicago, and that operators may enjoy getting a written record of their Dublin to determine transfers and linked-trip origins and performance--for the first time operators had written evi- destinations by tracking where a fare ID is next used to enter dence to show their family what a good job they were doing the system (18, 19, and unpublished work). This area is prom- in staying on schedule. ising for future research. 4.10 Higher Level Analyses 4.10.3 Headways and Other Measures on Shared Routes This section discusses analyses of AVL-APC data that cover extended periods of time or multiple routes. Many transit networks have trunks served by multiple lines or multiple patterns (branches) of a line. Some measures of activity on a trunk are simple aggregations of stop-level meas- 4.10.1 Comparisons and Aggregations ures; examples are schedule adherence and passenger load. For By comparing results of analyses done over selected dates, these purposes, all that is needed is an interface allowing one AVL-APC data can be used in beforeafter studies or to ana- to select the appropriate set of stops and patterns. However, lyze operations during special events or weather conditions. headways on a shared route can only be determined by going Trends analysis can be seen as simply an extension of the back to original stop or timepoint records, including data from beforeafter study, but it suggests a need for storing higher all the trips serving the trunk, and linking them where their level summaries in a separate database. An example is a respective route joins and leaves the common trunk. This pro- monthly systemwide report on schedule adherence. A transit cedure demands a special data structure for a route trunk, agency might specify measures that it wants to follow over something developed as part of this project (see Section 11.5). time, calculate those measures periodically (e.g., every month) from the detailed AVL-APC data archive, and save those period 4.10.4 Geographic Analyses summaries in a smaller, higher level database where they can be used for trends analysis. Transit agencies often want to do route-independent analy- Many analyses that involve aggregation or comparison over ses based on geography, including both demand analysis (how routes can benefit from AVL-APC data. One example is a many boardings occur in a certain area) and service quality periodic route performance comparison, which may include analysis (what is the on-time performance in a certain area). items such as on-time performance or total boardings along Integrating AVL-APC data with GIS models requires data with data from other sources such as scheduled vehicle-hours structures that link geographic locations to stops and route or farebox revenue. Another example is making annual sys- segments, and a process to extract and aggregate results for temwide passenger-miles estimates for reporting to the NTD, the selected stops and segments. which can be made by aggregating mean passenger-miles per For demand modeling, methods are needed to convert on- trip over all the trips in the schedule (see further discussion off counts at stops into trip generation rates in small traffic in Chapter 9). analysis zones. However, this specialized procedure could be These applications suggest having an automated process of driven equally by manual or automatically collected data; the periodically calculating and storing summary measures in challenge for APC data analysis is to export demand rates by higher level tables. stop for only a selected period of the day.