Click for next page ( 61


The National Academies | 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 60
Chapter 3. Schedule Building Level . Intermediate Schedule Building 3.3 Working on a More Intricate Schedule Even if you are new to Tip scheduling, do not be intimidated by the schedule and all Up to this point, new or trial schedules have relied on policy headways and calcu- of its detail. Even the most complex LEVEL lated running times for schedule development and construction. Headways and schedule can be understood if you 2B running times are the basic building blocks of all schedules. While they are some- times given to the scheduler, as in the previous exercises, it is often the responsi- review it one portion at a time. bility of the scheduler to build and adjust schedules based on an analysis of pas- senger load data and actual running time information collected by traffic checkers. As you continue to work your way through this manual, the examples become more complex and introduce new elements of scheduling. In this section, you will learn how to identify po- tential schedule changes based on variations in passenger load and running times. You will also traffic checkers learn to address the need for different running times throughout the day, and how to "smooth" Individuals who conduct ride checks the transition between periods of the day with different running times. Then in the advanced or point checks to collect ridership section, put these changes into practice by developing a revised schedule. and time-related data. The schedule in this exercise, for Route ROUTE 96--PASCO AVENUE --Pasco Avenue, is typical of many found peak-to-base ratio on medium to large transit systems. It has oa dw ay The ratio between the number of Br headways which vary, both by time of day Pa buses or trains required to operate sc Time point o Ru and direction. It contains running times that the schedule during the higher of gb y Ci n also change throughout the operating day. to the peak periods and by the num- rc is le i ll W It features multiple service patterns and a ber of buses in service in the "base" larger peak-to-base ratio, meaning that it period between the peaks. A peak Pa sc o to base ratio of 2.0 means that e vr has a significant peak service that is about Ha % greater than the amount of service twice as many buses are required operated during the base period between to operate peak period service as the peaks. Service then tapers off during the Garage midday service. The peak-to-base evening period and into late night. A map ratio greatly influences the runcut and the headway sheet for weekday service Es se x in terms of the number of straight and split runs that are possible. A Sh Sa on Route are shown below. elb nd y higher ratio means more split runs. Pasco Po t no in Mi t 96A Li bb y am hr is 96 W 3-30

OCR for page 60
Level . Intermediate Schedule Building Chapter 3. Schedule Building S C HE D UL E S HE E T R o u te 96 IN EF F : S eptem ber 15, 2007 M O N D A Y T H R U F R ID A Y P art 1 N O R TH B O U N D S O U TH B O U N D OUT Libby Sand Pasco Pasco Pasco Rugby Rugby Pasco Pasco Pasco Sand Libby IN BLK GAR Wishrm Point Essex Havre Willis Circle Circle Willis Havre Essex Point Wishrm GAR LVE. 5 4:10 4:20 4:28 4:35 4:43 4:53 5:05 1 3:45 4:05 4:18 4:26 4:33 4:41 4:50 4:58 5:05 5:13 5:23 5:35 6 4:55 5:05 5:13 5:20 5:28 5:37 5:49 3 4:05 4:35 4:48 4:56 5:03 5:11 5:20 5:29 5:37 5:47 5:57 6:05 7 4:39 4:59 5:08 5:16 5:23 5:31 5:35 5:44 5:52 6:02 6:12 6:19 5 5:05 5:19 5:29 5:37 5:46 5:50 5:59 6:07 6:17 6:27 6:45 9 4:59 5:19 5:29 5:39 5:47 5:56 6:05 6:14 6:22 6:32 6:42 6:57 1 5:35 5:49 5:59 6:07 6:16 6:20 6:29 6:37 6:47 6:57 7:05 6 5:49 5:59 6:09 6:17 6:26 6:35 6:44 6:52 7:02 7:12 7:17 3 6:05 6:19 6:29 6:37 6:46 6:50 7:01 7:10 7:21 7:32 7:45 7 6:19 6:29 6:39 6:47 6:56 7:05 7:16 7:25 7:36 7:46 7:57 8 5:55 6:25 6:39 6:50 6:59 7:10 7:15 7:26 7:35 7:46 7:57 8:05 4 6:17 6:37 6:49 7:00 7:09 7:20 7:25 7:36 7:45 7:56 8:06 8:23 5 6:45 6:59 7:10 7:19 7:30 7:35 7:46 7:55 8:06 8:17 8:35 9 6:57 7:09 7:20 7:29 7:40 7:45 7:56 8:05 8:16 8:26 8:46 1 7:05 7:19 7:30 7:39 7:50 7:55 8:06 8:15 8:26 8:37 9:05 6 7:17 7:29 7:40 7:49 8:00 8:05 8:16 8:25 8:36 8:46 8:53 10 6:55 7:25 7:39 7:50 7:59 8:10 8:15 8:26 8:35 8:46 8:57 9:17 2 7:17 7:37 7:49 8:00 8:09 8:20 8:25 8:36 8:45 8:56 9:06 9:23 3 7:45 7:59 8:10 8:19 8:30 8:35 8:46 8:55 9:06 9:17 9:35 7 7:57 8:09 8:20 8:29 8:40 8:45 8:56 9:05 9:16 9:26 9:46 8 8:05 8:19 8:30 8:39 8:50 8:55 9:06 9:15 9:26 9:37 9:57 4 8:23 8:34 8:44 8:52 9:02 9:07 9:18 9:27 9:38 9:48 9:53 5 8:35 8:48 8:58 9:06 9:16 9:22 9:31 9:38 9:48 9:59 10:05 6 8:53 9:04 9:14 9:21 9:31 9:37 9:46 9:53 10:03 10:13 10:23 1 9:05 9:18 9:28 9:35 9:44 9:52 10:01 10:08 10:18 10:29 10:35 2 9:23 9:33 9:43 9:50 9:59 10:07 10:16 10:23 10:33 10:43 10:53 3 9:35 9:48 9:58 10:05 10:14 10:22 10:31 10:38 10:48 10:59 11:05 4 9:53 10:03 10:13 10:20 10:29 10:37 10:46 10:53 11:03 11:13 11:23 5 10:05 10:18 10:28 10:35 10:44 10:52 11:01 11:08 11:18 11:29 11:35 6 10:23 10:33 10:43 10:50 10:59 11:07 11:16 11:23 11:33 11:43 11:53 1 10:35 10:48 10:58 11:05 11:14 11:22 11:31 11:38 11:48 11:59 12:05 2 10:53 11:03 11:13 11:20 11:29 11:37 11:46 11:53 12:03 12:13 12:23 3 11:05 11:18 11:28 11:35 11:44 11:52 12:01 12:08 12:18 12:29 12:35 4 11:23 11:33 11:43 11:50 11:59 12:07 12:16 12:23 12:33 12:43 12:53 5 11:35 11:48 11:58 12:05 12:14 12:22 12:31 12:38 12:48 12:59 13:05 6 11:53 12:03 12:13 12:20 12:29 12:37 12:46 12:53 13:03 13:13 13:23 1 12:05 12:18 12:28 12:35 12:44 12:52 13:01 13:08 13:18 13:29 13:35 2 12:23 12:33 12:43 12:50 12:59 13:07 13:16 13:23 13:33 13:43 13:53 3 12:35 12:48 12:58 13:05 13:14 13:22 13:31 13:38 13:48 13:59 14:05 4 12:53 13:03 13:13 13:20 13:29 13:35 13:45 13:52 14:03 14:13 14:23 5 13:05 13:18 13:28 13:35 13:44 13:50 14:00 14:07 14:18 14:29 14:35 6 13:23 13:33 13:43 13:50 13:59 14:05 14:15 14:22 14:33 14:43 14:53 3-31

OCR for page 60
Chapter 3. Schedule Building Level . Intermediate Schedule Building S C HE D UL E S HE E T R o u te 96 IN EF F : S eptem ber 15, 2007 M O N D A Y T H R U F R ID A Y P art 2 N O R TH B O U N D S O U TH B O U N D OUT Libby Sand Pasco Pasco Pasco Rugby Rugby Pasco Pasco Pasco Sand Libby IN BLK GAR Wishrm Point Essex Havre Willis Circle Circle Willis Havre Essex Point Wishrm GAR LVE. 1 13:35 13:48 13:59 14:06 14:16 14:20 14:30 14:37 14:48 14:59 15:05 2 13:53 14:03 14:14 14:21 14:31 14:35 14:45 14:52 15:03 15:13 15:23 3 14:05 14:18 14:29 14:36 14:46 14:50 15:00 15:07 15:18 15:29 15:35 4 14:23 14:33 14:44 14:51 15:01 15:05 15:15 15:22 15:33 15:43 15:48 5 14:35 14:48 14:59 15:06 15:16 15:20 15:30 15:37 15:48 15:59 16:05 12 15:23 15:33 15:43 15:50 16:01 16:11 16:15 6 14:53 15:03 15:14 15:21 15:31 15:45 15:55 16:02 16:13 16:24 16:32 1 15:05 15:18 15:29 15:36 15:46 15:57 16:07 16:14 16:25 16:35 16:45 2 15:23 15:33 15:44 15:51 16:01 16:09 16:19 16:26 16:37 16:48 16:52 3 15:35 15:48 15:59 16:06 16:16 16:20 16:31 16:39 16:50 17:01 17:05 4 15:48 15:58 16:09 16:16 16:26 16:30 16:41 16:49 17:00 17:12 17:15 11 15:53 16:08 16:19 16:26 16:36 16:40 16:51 16:59 17:10 17:21 17:28 5 16:05 16:18 16:29 16:36 16:46 16:50 17:01 17:09 17:20 17:32 17:36 12 16:15 16:25 16:36 16:44 16:55 17:00 17:11 17:19 17:30 17:41 17:53 13 16:20 16:35 16:46 16:54 17:05 17:10 17:21 17:29 17:40 17:52 18:05 6 16:32 16:45 16:56 17:04 17:15 17:20 17:31 17:39 17:50 18:01 18:21 1 16:45 16:55 17:06 17:14 17:25 17:30 17:40 17:48 17:58 18:10 18:35 2 16:52 17:05 17:16 17:24 17:35 17:40 17:50 17:58 18:08 18:19 18:23 3 17:05 17:15 17:26 17:34 17:45 17:52 18:02 18:10 18:20 18:32 18:52 4 17:15 17:28 17:39 17:47 17:58 18:05 18:15 18:23 18:33 18:44 18:53 11 17:28 17:38 17:49 17:57 18:08 18:18 5 17:36 17:49 17:59 18:06 18:15 18:20 18:30 18:38 18:48 19:00 19:05 12 17:53 18:03 18:13 18:20 18:29 18:35 18:45 18:53 19:03 19:14 19:23 13 18:05 18:18 18:28 18:35 18:44 18:50 18:59 19:06 19:15 19:26 19:35 2 18:23 18:33 18:43 18:50 18:59 19:05 19:14 19:21 19:30 19:40 19:53 1 18:35 18:48 18:58 19:05 19:14 19:25 19:34 19:41 19:50 20:01 20:06 4 18:53 19:03 19:13 19:20 19:29 19:45 19:54 20:01 20:10 20:20 20:28 5 19:05 19:18 19:28 19:35 19:44 19:54 12 19:23 19:33 19:43 19:50 19:59 20:05 20:14 20:21 20:30 20:41 20:46 13 19:35 19:48 19:58 20:05 20:14 20:25 20:34 20:41 20:50 21:00 21:08 2 19:53 20:03 20:12 20:19 20:26 20:36 1 20:06 20:18 20:26 20:32 20:39 20:45 20:54 21:01 21:10 21:21 21:26 4 20:28 20:38 20:46 20:52 20:59 21:05 21:12 21:18 21:26 21:35 21:48 12 20:46 20:58 21:06 21:12 21:19 21:25 21:32 21:38 21:46 21:56 22:06 13 21:08 21:18 21:26 21:32 21:39 21:45 21:52 21:58 22:06 22:15 22:28 1 21:26 21:38 21:46 21:52 21:59 22:05 22:12 22:18 22:26 22:36 22:46 4 21:48 21:58 22:06 22:12 22:19 22:25 22:32 22:38 22:46 22:55 23:08 12 22:06 22:18 22:26 22:32 22:39 22:45 22:52 22:58 23:06 23:16 23:26 13 22:28 22:38 22:46 22:52 22:59 23:05 23:12 23:18 23:26 23:35 23:48 1 22:46 22:58 23:06 23:12 23:19 23:25 23:32 23:38 23:46 23:56 0:06 4 23:08 23:18 23:26 23:32 23:39 23:49 12 23:26 23:38 23:46 23:52 23:59 0:05 0:12 0:18 0:26 0:36 0:56 13 23:48 23:58 0:06 0:12 0:19 0:29 1 0:06 0:18 0:26 0:32 0:39 0:49 3-32

OCR for page 60
Level . Intermediate Schedule Building Chapter 3. Schedule Building The most productive Things you need to know for this exercise: Tip transit routes are those . Ridership on Route has been growing, as first revealed by daily farebox passenger that have balanced demand in boarding figures. both directions. Routes with a very pronounced peak direction may . The route has also developed a problem with keeping on schedule. The early warning be very full in one direction, but for this is usually declining on-time performance figures. An even more accurate warn- must return with very light loads to ing may come from the drivers, who let management know they are having trouble recycle. Designing routes to serve keeping the schedule during certain times of the day due to insufficient running time important nodes on both ends will and heavy boardings at several locations. generally improve productivity. The maximum load point (max point for short) has been determined over time for this route from ride checks (see the discussion in Chapter , Inputs to the Scheduling Process, about ride check analysis). It is at a major intersection, Pasco & Havre, which is also a time point and a fairly sizable transfer point. Route has changed character over the years, from a route with a very pronounced peak direction to one where ridership is generally well balanced in both directions during peak periods. This trend has been beneficial in making better use of sched- uled trips during peaks: previously non-peak-direction trips operated with few passengers but now trips carry good loads in both directions during the peak period. The route has also seen a reduction in peak loads in the peak direction, allowing a slightly wider headway to carry similar numbers of riders per trip. Analysis of trips by time periods provides a sense of how well demand (riders) meets supply (seats plus standees). Schedule departments have traditionally used analysis periods of - to -minute intervals during times when headways are minutes or less and - to -minute intervals during times when headways are minutes or more. The load average spreadsheet for Route is shown on the following page. The load average spreadsheet has thicker lines at hourly intervals during the base and half hourly intervals during the two peaks. We use the scheduled arrival time at the max point (the "due" column, second from the left) to determine where trips fit in the time breaks. The load analysis examines each time interval. The spreadsheet also contains formulas in each time interval: the first sums all passengers observed (this check collected the maximum number of passengers, whether that was the arriving load or the leaving load); the second calculates the average load per trip within the time interval. Average load per trip within the time interval is shown in bold on the load average spreadsheet. 3-33

OCR for page 60
Chapter 3. Schedule Building Level . Intermediate Schedule Building S A M P L E TR A N S IT A G E N C Y AVE R AG E D M AX LO AD P O IN T C H E C K S R O U T E 96 - P A S C O A V EN U E W EEK D A Y S C H ED N O . 5 S O UTHBO UND P AS CO & HAV RE A R R IV IN G L O A D Tues day A pril 1 Thurs day A pril 10 W ednes day M ay 7 S unny 47 degrees C loudy , rain, 54 degrees O verc as t, c ool, 57 degrees BLK DUE ACTUAL LO AD T o t/A vg ACTUAL LO AD T o t/A vg A C T U A L LO AD T o t/A vg A V ER A G E 5 6:07 6:08 11 6:09 14 6:06 12 9 6:22 6:23 27 6:22 30 6:22 26 1 6:37 6:37 22 88 6:38 24 97 6:39 23 87 6 6:52 6:51 28 22 6:55 29 24 6:53 26 22 23 3 7:10 7:10 33 69 7:09 35 70 7:12 31 64 7 7:25 7:26 36 35 7:26 35 35 7:26 33 32 34 8 7:35 7:36 38 7:38 37 7:37 39 4 7:45 7:42 41 123 7:46 43 126 7:46 40 121 5 7:55 7:58 44 41 7:58 46 42 7:58 42 40 41 9 8:05 8:07 40 8:05 43 8:06 41 1 8:15 8:16 42 120 8:16 43 126 8:15 40 117 6 8:25 8:27 38 40 8:28 40 42 8:24 36 39 40 10 8:35 8:35 36 8:35 37 8:37 33 2 8:45 8:46 32 102 8:45 35 104 8:46 33 96 3 7 8:55 9:05 8:57 9:04 34 27 34 8:56 9:06 32 33 35 8:55 9:05 30 30 32 34 peak of the peak 8 9:15 9:26 40 9:15 27 9:16 25 4 5 9:27 9:38 9:27 9:41 14 23 130 9:28 9:40 26 24 132 9:28 9:41 24 21 124 The absolute busiest time interval 6 1 9:53 10:08 9:54 10:07 26 21 26 9:55 10:08 22 23 26 9:53 10:08 24 22 25 26 (measured in short increments such 2 3 10:23 10:38 10:23 10:37 19 22 80 10:24 10:37 22 23 90 10:25 10:40 20 23 83 as 15 or 30 minutes, depending on 4 5 10:53 11:08 10:55 11:08 18 24 20 10:53 11:10 22 29 23 10:54 11:08 18 27 21 21 headway) during the peak period, 6 1 11:23 11:38 11:25 11:37 26 25 101 11:23 11:39 24 25 106 11:23 11:38 25 24 102 in terms of passenger demand and 2 11:53 11:53 26 25 11:58 28 27 11:55 26 26 26 service. 3 4 12:08 12:23 12:08 12:24 23 25 12:09 12:20 25 15 12:09 12:24 26 27 Peaks of the peak loads 5 12:38 12:40 24 100 12:38 36 105 12:38 23 104 6 12:53 12:54 28 25 12:54 29 26 12:53 28 26 26 1 13:08 13:08 28 13:08 30 13:08 29 2 13:23 13:23 27 13:24 29 13:24 26 3 13:38 13:39 32 106 13:39 31 115 13:39 29 106 4 13:52 13:50 19 27 13:53 25 29 13:54 22 27 27 5 14:07 14:07 33 14:08 35 14:09 32 6 14:22 14:24 28 14:22 32 14:22 28 1 14:37 14:34 22 110 14:37 33 136 14:40 29 121 2 14:52 14:50 27 28 14:55 36 34 14:53 32 30 31 3 15:07 15:07 29 15:07 31 15:06 30 4 15:22 15:22 28 15:22 30 15:23 29 5 15:37 15:39 31 120 15:36 32 127 15:37 27 116 14 15:50 15:53 32 30 15:50 34 32 15:54 30 29 30 6 16:02 16:01 36 16:04 38 16:04 35 1 16:14 16:15 34 109 16:15 37 116 16:14 29 97 2 16:26 16:28 39 36 16:28 41 39 16:27 33 32 36 3 16:39 16:41 38 16:38 40 16:41 42 4 16:49 16:49 43 128 16:50 46 135 16:51 44 136 12 16:59 17:01 47 43 17:00 49 45 17:00 50 45 44 5 17:09 17:13 51 17:10 55 17:09 54 13 17:19 17:22 53 153 17:21 58 166 17:20 56 164 14 17:29 17:32 49 51 17:31 53 55 17:33 54 55 54 6 17:39 17:43 44 17:42 48 17:41 49 1 17:48 17:48 39 118 17:48 40 130 17:47 41 127 11 17:58 17:58 35 39 17:58 42 43 17:58 37 42 42 3 18:10 18:11 30 18:10 37 18:11 33 2 18:23 18:24 27 18:25 32 18:25 30 4 18:38 18:39 29 110 18:39 30 127 18:39 26 118 13 18:53 18:53 24 28 18:53 28 32 18:54 29 30 30 14 19:06 19:07 20 19:09 28 19:06 24 6 19:21 19:20 18 53 19:22 20 68 19:22 19 60 1 19:41 19:41 15 18 19:42 20 23 19:41 17 20 20 2 20:01 20:02 19 20:01 22 20:02 20 13 20:21 20:21 12 47 20:23 16 56 20:21 15 54 14 20:41 20:40 16 16 20:40 18 19 20:43 19 18 17 1 21:01 21:01 11 21:01 15 21:00 12 2 21:18 21:19 9 21:18 12 21:18 14 13 21:38 21:39 14 44 21:42 13 51 21:39 11 49 14 22:01 22:01 10 11 21:59 11 13 21:59 12 12 12 3-34

OCR for page 60
Level . Intermediate Schedule Building Chapter 3. Schedule Building Even if you have the Route operates past the maximum load point every minutes during the base period and Tip resources for trip-by-trip every during the peaks. So the peak periods are divided into -minute intervals and the analysis, traffic check data should base period into -minute intervals. be averaged over a time interval of three to four times the headway. The time interval is critical. We do not want to miss the high peak of the peak loads by averag- Loading on individual trips varies ing these with too many trips on either side (in the shoulders of the peak). We cannot always greatly, while loading over a time schedule effectively for the peak of the peak demand, but averaging trips that are too far away period remains fairly constant. from the peak of the peak can give us a false lower estimation of peak demand. Use of a - minute interval during the peak periods avoids this pitfall on Route . Keep in mind the overall purpose, which is to be able to compare observed loads against es- tablished service standards. Looking at individual trip loadings is too precise for analysis when service is this frequent, as trips can run erratically due to traffic and other factors largely out of the control of the transit system. These factors can produce overloads on one trip while sur- rounding trips are running with fewer passengers than normal. Bus operators also vary as to style of driving. Some are faster than others, meaning they are able to handle the business of collecting fares, handing out information, tying down wheelchairs, and maneuvering through traffic better than their fellow operators. The averaging process within the analysis period helps to compensate for these differences. The average load spreadsheet is laid out to accommodate three separate days' checks. We could add or subtract columns based on the number of days of data available. As always, more than one day of data is recommended for any schedule analysis. For this example, we concentrate on the maximum load point of the route for data collection. Route has two branches, designated and A on the driver's route map, although A may not appear on public timetables or bus headsigns. A thorough review of ridership would also call for a point check at the junction of these two branches, at the time point of Pasco & Essex. The existing schedule is built with all trips alternating between each branch throughout the day. This is the most practical arrangement when route branches exist, but it does not mean that ridership justifies this equal distribution of service. One branch could be significantly weaker and only justify, say, one out of three trips. Since the balance of ridership between the two branches usually does not change much over time, it is not necessary to schedule point checks at this location on a regular basis. For the purpose of this exercise, ridership on each branch is assumed to be roughly equal and checks made at Pasco & Essex have shown very little shift over the past several years. 1 Schedules material will often contain notations for operating personnel--including telephone information for customer service representatives--that do not appear on written public materials. Identifying the branch of a route (such as the 96A in our example) for operational purposes, but not for the public, is not uncommon. 3-35

OCR for page 60
Chapter 3. Schedule Building Level . Intermediate Schedule Building The amount of data With all load numbers laid out conveniently on this sheet, it is easy to see the dimensions of Tip the peak. During the AM peak in the southbound direction, service levels are adequate. On available for analysis will depend both on the average, there are two to three standees per trip between : and : AM (average load is number of days of checking and on ) and between : and : AM (average load is ). We would then want to look at the the schedule used to deploy traffic northbound direction loads, the known peak direction, before making any judgment on service checkers. Traffic checking shifts adjustments. can be scheduled in different ways. A typical checking assignment will In the PM, it is a different situation. For the : - : peak half hour, the average load is , cover a location for 16 straight five higher than the standard allows. The time periods before and after this period show aver- hours with two shifts, often 5:00 ages of and , respectively, which are within the standard. An additional trip is needed in AM to 9:00 PM or 6:00 AM to 10:00 this half-hour period in order to bring the loads down to within the maximum specified in the PM. These shifts cover the times standard ( ). A simple calculation shows that a fourth trip during this half hour would bring when there is usually the great- the average load down to slightly more than . Since the neighboring time intervals could est need to observe loads and trip benefit from the added trip, we would extend the reduced headway for more than just the performance. At times there may peak half-hour period. be a need to look at the ends of the service day. In this case, a "start to It is a good idea to sketch any potential change to see what the proposed schedule might look finish" schedule, which will typically like. Here is an initial proposal for fixing the overload period: leave a portion of the middle of the day uncovered, can be used. For Southbound times at the max point (Pasco & Havre): Route 96, the early checker would be in place at Pasco & Havre at Present Proposed about 4:25 AM and remain there 16:39 16:39 until 12:25 PM. The afternoon 16:49 16:48 checker would be assigned there at least by 4:30 PM and remain there 16:59 16:56 until the last trip, a northbound 17:04 bus, passes at 12:26 AM. Combin- 17:09 17:12 ing both of these types of checks together provides the whole picture 17:19 17:20 on max point ridership. 17:29 17:28 17:39 17:37 One final point on scheduling traffic checks is that checkers, like all em- 17:48 17:47 ployees, will need breaks for meal times and rest periods. Check- ing over multiple days can allow This particular adjustment spreads the benefit of the additional trip over an approximate checker breaks to occur at different - minute period, which helps to lower the peak count on six trips. What other effects does times on different days, ensuring this have on the schedule? The scheduler knows that the layover at both ends of the route is that there are no significant gaps in already tight, so it is unlikely that this extra trip can be provided using the existing number of data collection. buses. But it is too early in the process to finalize this change. The scheduler also noted that 3-36

OCR for page 60
Level . Intermediate Schedule Building Chapter 3. Schedule Building passengers most of the trips in the peak, as well as during other times in the operating day, seem to be ar- per-minute riving late. A running time analysis is appropriate before crafting a final change proposal. or PPM for short, is the measure- Before analyzing the running time, the scheduler needs to examine the base and evening ment of how many people accumu- periods. Different standards are provided for those periods. During the base, the agency's late every minute at all bus stops standards call for a minimum % load during the midday period and all day on Saturday. This waiting for service in the direction guideline is meant to allow for a better base service frequency, which is expected to attract being analyzed. It is generally expressed as a decimal number, more riders. That translates to a half seated load of at least on each trip. The hourly time such as 5.8 PPM. What that means analyses between : AM and : PM ( : ) all exceed that, ranging from to a high of is that, between the time the last just before the shoulder of the PM peak. The scheduler could actually look at the effect of add- trip departed and the next trip is at ing some base service. The passengers per-minute (PPM) calculation helps here, providing a the stop, 5.8 potential customers quick way of performing a what-if analysis. arrive at this bus stop every minute. That number helps tell us how How is the PPM figure calculated? The following example using the average loads of the three much service we need to run to ridechecks on Route illustrates the calculation during the peak of the peak just after : meet the demand within our service PM. standard. Trip Minutes since Minutes since Arrive Load last trip Trip PPM 16:49 Period PPM 16:49 0 0 16:59 49 10 4.90 10 4.90 17:09 53 10 5.30 20 5.10 17:19 56 10 5.60 30 5.27 17:29 52 10 5.20 40 5.23 Total 214 How many passengers should be on each bus? If the service standards call for a seated load plus a % standee rate during the peak period (a common standard) and the buses assigned to this route seat , then each trip should be scheduled to carry a maximum of passengers past the max point. In this example, all four trips exceed a load of . At . PPM for the : trip, if the bus were only one minute earlier, there would be instead of patrons on board. Since the PPM can be highly volatile, especially when the service is running erratically, it is easier to "smooth" out the calculation by looking at a number of trips over time. For the entire 3-37

OCR for page 60
Chapter 3. Schedule Building Level . Intermediate Schedule Building -minute time period we have analyzed, the PPM is . . Divided into a minute cycle time, we find that a nine-minute headway would provide the service level we are looking for. It is also useful if we decide to adjust loads on our sheet to account for early and late trips. To keep our discussion of max point analysis a bit simpler, we will "eyeball" the early or late trips, their corresponding loads, and mentally make adjustments as we record averages by time period. This practice is certainly acceptable provided that we know our route and a little about what causes early or late operation. If the max load numbers are accurate (and they are by this point), then our judgment should not cause us to err significantly in producing a representative average for the time period. We can take the noon hour, where the average load is and calculate what would happen if we went from four to five buses per hour (a -minute headway on the trunk portion of the route). We have a three-day average of passengers over a -minute time period, which yields a PPM of . . Multiplying that number by (the new headway) gives us , the num- ber of average passengers on-board for the improved service. Since our goal is to have no fewer than on board, a -minute headway would meet our goal. We know we need to add PM peak service and we could also add some service in the base, but the base addition is discretionary. The current average load of represents only % of the seats filled and no standees. The following considerations will influence the decision: Can the agency afford the additional - platform hours that one or two added buses would cost us? Does the ridership trend in the base on this route continue to show growth? Is the average load in any of the time periods approaching a seated load? Is the peak-to-base ratio on this route high enough to justify more base service in order to help produce a more economical runcut? The last two points are the most germane in helping to make the decision. The average loads are not high enough to be a concern at this point and the peak-to-base ratio is still reasonable (six base buses compared with during the peak, giving a ratio of . ). Anything below a ratio of . will produce a reasonably cost-efficient runcut in most situations. (More on what makes a runcut cost-efficient is presented in Chapter .) The decision, then, is to adjust the PM peak service to address the overloads and leave the base period alone for now. The next step is analyzing running time. 3-38

OCR for page 60
Level . Intermediate Schedule Building Chapter 3. Schedule Building When planning peak Running Time Checks Tip hour service, it is often Running time is even more volatile than passenger loads since it depends on street conditions, useful to have standards that reflect and individual driver habits. To account for variations in running time, schedulers should err the maximum number of people on the side of having more checks available to analyze. Ride checks can be taken continuously that should be on board the bus at the maximum load point. Higher throughout the system and, like point checks, can be supplemented with regular discussions loads will result in increased ser- with drivers and other operating personnel who are known to give accurate feedback. vice. For off-peak service, includ- The early warning mechanism in this example is noting that a number of the trips on Route ing evenings and weekends, it is showed up late at the maximum load point. Also, there were no more than a handful of early useful to have both maximum load trips, which is an indicator that running time sufficiency needs to be reviewed. For this pur- standards and standards that reflect pose, the scheduler once again turns to a three-day check summary. The most efficient way of the minimum number of passengers obtaining a multi-day summary of running time data is from automated data collection meth- that should be on board before ods, such as APCs or an AVL system. But not all transit properties have APCs or have plans to service is reduced. procure them in the near future. What are the best manual methods of spreading and calculating running times? There are two basic steps in this task. First is the analysis of individual checks. This includes a review of the check itself for normalcy of operation and for possible hand adjustment to account for any slightly erratic operation so that the times will reflect what would be considered the "normal case" if the aberrations were removed. Once the "normal" or "representative" checks are accumulated and the "outlier" checks have been discarded, the second step is to post the times to a comparison sheet where the multiple days' checks can be readily studied and averaged. The example on page - shows a portion of an expanded headway sheet that could be marked up manually or filled in electronically, complete with formulas to average the several days' observed running times. These times have been analyzed beforehand, so no additional notations or allowances need to be made. The spreadsheet has two columns beside the observed running time totals that show the pres- ent running time period breaks and the scheduled running time. The bold horizontal lines mark the beginning and end of periods for which running times will be averaged. Three-day averag- es of running times for each trip in each direction are shown in the columns to the right of the northbound and southbound schedules. Next, the recommended (or "decided") running times for a given time period are displayed. 3-39

OCR for page 60
Chapter 3. Schedule Building Level . Intermediate Schedule Building Establishing Running Time Periods There are two separate steps in analyzing running time: . Break the operating day down into individual running time periods for each direction . Analyze and adjust the individual running times between time points so they add up to the total for the one-way trip. Running time periods, also called "classes," "breaks," and "levels" at various systems, are times of the day when a specific running time is in effect. The number for each schedule can vary from one running time period for the entire day to or more. They are established or adjusted by looking at trends in how total trip running times vary over the day. Total one-way running times have traditionally been used to find where these breaks occur, as the individual times between time points provide too much information to digest and consider manually. Comput- erized scheduling packages simplify the adjustment of running times by segment. We begin by looking at the total running time for each trip in each direction. We are trying to spot trends. Running times tend to jump around a bit. In the example below, note that even with the variances you can see when the trend is moving upward or downward. 3-40

OCR for page 60
Level . Intermediate Schedule Building Chapter 3. Schedule Building Trip Total Time 5:06 42 5:26 44 5:40 43 5:55 46 6:10 49 6:20 48 6:20 51 6:40 49 6:50 50 7:00 53 7:10 55 7:20 54 7:30 56 7:40 55 7:50 54 8:00 55 8:10 53 8:20 51 8:35 49 8:50 48 9:05 50 9:25 47 9:45 46 10:05 47 The figures represent total trip times during the early morning, AM peak, and mid morning base periods. On first glance, it looks as if there is no order at all to totals, much less a trend. But experience suggests that even with some slight variations among adjacent trips, the trend should be for increases during the buildup to the : AM peak-of-the-peak and decreases afterward. 3-41

OCR for page 60
Chapter 3. Schedule Building Level . Intermediate Schedule Building The first three trips are all within a minute of an average minutes of running time. The running time for the fourth trip at : is notably higher, so a time break before the : trip is appropriate. Likewise, there is a similarity in running times between : and : . The times range between and minutes and average . minutes. Toward the end of the table in the base period, running time is in the to minute range. The average of the last three trips on our table is . Thus, we have dealt with the easier-to-spot areas first. Between the early trips and the peak and between the peak and the base are transition areas. Times either ramp up or down relatively quickly. The question is: how big is the gap between the highest peak number and the lowest early or base number? Can one transitional running time bridge the gap or is the difference large enough to require two or more running time periods? Following is the same table of trip times sliced into different running time periods using the method just described. 3-42

OCR for page 60
Level . Intermediate Schedule Building Chapter 3. Schedule Building Trip Total Time Average 5:06 42 5:26 44 43 5:40 43 5:55 46 6:10 49 6:20 48 6:20 51 48.8 6:40 49 6:50 50 7:00 53 7:10 55 7:20 54 7:30 56 54.4 7:40 55 7:50 54 8:00 55 8:10 53 8:20 51 8:35 49 49.5 8:50 48 9:05 50 9:25 47 9:45 46 47 10:05 47 The transition period from the early AM to the AM peak must help bridge an increase of minutes. Analysis of the actual trip times shows a range from to minutes during the tran- sition period between : and : . The average of all six trips is . rounded to , about halfway between the adjacent averages. In this case, the decision is to go with the and just one transition running time period. 3-43

OCR for page 60
Chapter 3. Schedule Building Level . Intermediate Schedule Building An alternate but equally reasonable choice would have been to divide the period into two three-trip periods. Under this option, the first three trips average , while the second three average . The result would be two transitional running time periods and a better smoothing (in terms of less headway variation) of the running time transitions. Either one or two transition running time periods is acceptable, so long as the scheduler plans to perform manual smooth- ing of the trips to keep headways as close to minutes as possible. The second transition, into the base, is less dramatic, having a difference of only seven min- utes. One transitional running time period should suffice here and the times average , which is just about halfway between the two times. A similar effort, not shown here, is needed going into and out of the PM peak. From the PM peak into the night running time, two transition periods might be needed because of the large drop in running time. This same principle applies to our large running time summary sheet for Route . The first step is to "eyeball" the totals on the sheet, looking especially at the peaks and valleys during the day. Then group similar running time totals into periods. Laying out individual running time observations this way provides an ideal way of looking at trends. As noted earlier, the spread- sheet has two columns beside the observed running time totals that show the present run- ning time period breaks and the scheduled running time. Working on the computer or on hard copies of these sheets, the scheduler can make notations about any needed changes , both to the time when running time periods change and to new, proposed totals. Note again that the scheduler works first with totals and then works backwards to adjust the individual running times. The analysis of the AM peak generally shows the present running times are adequate. Individu- al times on trips and trip segments do vary, but over the full period of the peak, actual running times are very close to scheduled running times for each branch in both the northbound and southbound directions. The same holds true of the base. 3-44

OCR for page 60
Level . Intermediate Schedule Building Chapter 3. Schedule Building HE A D W A Y S HE E T Line 96 IN EF F : S eptem ber 15, 2007 M O N D A Y T H R U F R ID A Y H E A D W A Y W ITH 3 -D A Y RUNN IN G TIM E C H E C K S A D D E D D IR E C TIO N = N O R TH - S O U TH AM Libby Sand Pasco Pasco Pasco Rugby Avg Avg Proposed Proposed Rugby Pasco Pasco Pasco Sand Libby IN Avg Avg Proposed Proposed Wishrm Point Essex Havre Willis Circle Lib/Wsh Sand Pt Lib/Wsh Sand Pt Circle Willis Havre Essex Point Wishrm GAR Lib/Wsh Sand Pt Sand Pt Lib/Wsh 5:05 5:19 5:29 5:37 5:46 5:50 5:59 6:07 6:17 6:27 5:19 5:29 5:39 5:47 5:56 6:05 6:14 6:22 6:32 6:42 9.5 8 9 10 10 7.5 9.5 9 9 7.5 10 9 9.5 7.67 9.5 9.33 36 5:35 5:49 5:59 6:07 6:16 6:20 6:29 6:37 6:47 6:57 37 37 9 8 10 9.5 8.5 8.5 9.5 9.5 9.5 8 11 10 9 8.17 10 .2 9.67 37 5:49 5:59 6:09 6:17 6:26 6:35 6:44 6:52 7:02 7:12 41 37 9.5 8 10 10 8.5 9 9.5 9.5 9.5 9 9.5 9.5 9.17 8.67 9.67 9.67 37 6:05 6:19 6:29 6:37 6:46 6:50 7:01 7:10 7:21 7:32 14 9.5 8 9.5 10 9 10 11 13 9.5 8.5 9.5 10 .5 9.5 11 11 .5 13 .5 10 8 9 10 .5 9.5 11 10 .5 13 .5 9.67 8.17 9.33 41 10 .3 9.33 10 .7 11 41 6:19 6:29 6:39 6:47 6:56 7:05 7:16 7:25 7:36 7:46 10 9.5 8.5 9 11 9 10 10 10 10 9 9.5 10 .5 8.5 11 9 9.5 10 .5 8.5 9.5 10 .5 9 11 .5 12 9.83 10 8.67 9.33 38 10 .7 8.83 10 .8 10 .3 41 6:25 6:39 6:50 6:59 7:10 7:15 7:26 7:35 7:46 7:57 13 12 9 10 11 .5 9 11 11 14 11 8 11 11 9 12 10 14 11 9 12 11 8 11 10 13 .7 11 .3 8.67 11 45 11 .2 8.67 11 .3 10 .3 42 6:37 6:49 7:00 7:09 7:20 7:25 7:36 7:45 7:56 8:06 11 12 9 11 11 9 11 10 41 42 12 11 9 12 12 9 10 10 12 12 8 11 11 10 11 11 11 .7 11 .7 8.67 11 .3 43 11 .3 9.33 10 .7 10 .3 42 6:45 6:59 7:10 7:19 7:30 7:35 7:46 7:55 8:06 8:17 14 11 9 12 11 9 12 11 14 11 10 11 11 9 11 12 15 11 9 11 12 10 11 11 14 .3 11 9.33 11 .3 46 11 .3 9.33 11 .3 11 .3 43 6:57 7:09 7:20 7:29 7:40 7:45 7:56 8:05 8:16 8:26 8:46 12 11 9 11 10 9 11 10 13 10 8 11 45 43 11 8 10 11 11 11 9 12 11 9 11 11 3-45

OCR for page 60
Chapter 3. Schedule Building Level . Intermediate Schedule Building A look at the PM peak shows the inadequacy noted earlier based on the analysis of the passen- ger load checks at the max point. More than a few trips arrived late at Pasco & Havre by one or two minutes. At the same time, overcrowding was seen on southbound trips during the height of the PM peak. These two trends suggest the need to rebuild the PM portion of the schedule. HEA D W A Y S HEE T L in e 96 IN EFF : S ep t e m be r 15 , 200 7 M O NDA Y T HRU F R IDA Y HE A D W A Y W ITH 3-D A Y R U NNING TIM E C HE C K S A D D E D D IR E C TIO N = N O R TH - S O U TH P M P E A K P E R IO D Libby Sand Pasco Pasco Pasco Rugby Avg Avg Proposed Proposed Rugby Pasco Pasco Pasco Sand Libby Avg Avg Proposed Proposed Wishrm Point Essex Havre Willis Circle Lib/Wsh Sand Pt Lib/Wsh Sand Pt Circle Willis Havre Essex Point Wishrm LVE. Lib/Wsh Sand Pt Sand Pt Lib/Wsh 15 : 3 5 15 : 4 8 15 : 5 9 16 : 0 6 16 : 1 6 16 : 2 0 16 : 3 1 16 : 3 9 16 : 5 0 17 : 0 1 17 : 0 5 13 12 7 10 12 9 12 12 12 11 7 11 12 9 11 12 13 12 7 10 13 8 12 13 12.7 11.7 7 10.3 42 12.3 8.67 11.7 12.3 45 15 : 4 8 15 : 5 8 16 : 0 9 16 : 1 6 16 : 2 6 16 : 3 0 16 : 4 1 16 : 4 9 17 : 0 0 17 : 1 2 17 : 1 5 11 11 7 10 13 9 13 13 45 46 10 11 8 9 12 8 12 13 11 10 8 10 11 9 12 13 10.7 10.7 7.67 9.67 39 12 8.67 12.3 13 46 16 : 0 8 16 : 1 9 16 : 2 6 16 : 3 6 16 : 4 0 16 : 5 1 16 : 5 9 17 : 1 0 17 : 2 1 17 : 2 8 12 7 10 12 9 12 12 11 7 11 13 9 12 13 11 8 10 12 8 11 13 11.3 7.33 10.3 29 12.3 8.67 11.7 12.7 45 16 : 0 5 16 : 1 8 16 : 2 9 16 : 3 6 16 : 4 6 16 : 5 0 17 : 0 1 17 : 0 9 17 : 2 0 17 : 3 2 17 : 3 6 13 12 7 10 12 8 12 13 13 11 8 10 12 9 13 14 14 12 8 10 11 8 12 13 13.3 11.7 7.67 10 43 11.7 8.33 12.3 13.3 46 16 : 1 5 16 : 2 5 16 : 3 6 16 : 4 4 16 : 5 5 17 : 0 0 17 : 1 1 17 : 1 9 17 : 3 0 17 : 4 1 17 : 5 3 11 12 8 10 12 8 12 12 11 12 8 11 45 42 11 9 13 12 10 11 9 11 13 9 12 11 10.7 11.7 8.33 10.7 41 12 8.67 12.3 11.7 45 1 6 :3 5 1 6 :4 6 1 6 :5 4 1 7 :0 5 17 : 1 0 17 : 2 1 17 : 2 9 17 : 4 0 17 : 5 2 18 : 0 5 12 8 11 12 9 12 13 12 8 10 12 8 11 13 45 46 12 7 11 13 8 12 14 12 7.67 10.7 30 12.3 8.33 11.7 13.3 46 16 : 3 2 16 : 4 5 16 : 5 6 17 : 0 4 17 : 1 5 17 : 2 0 17 : 3 1 17 : 3 9 17 : 5 0 18 : 0 1 14 12 8 11 12 9 11 13 13 11 7 12 13 8 12 12 13 11 7 11 12 8 12 11 13.3 11.3 7.33 11.3 43 12.3 8.33 11.7 12 44 16 : 4 5 16 : 5 5 17 : 0 6 17 : 1 4 17 : 2 5 17 : 3 0 17 : 4 0 17 : 4 8 17 : 5 8 18 : 1 0 18 : 3 5 11 11 7 11 45 42 12 8 11 13 11 12 8 11 12 9 12 12 11 12 8 10 11 8 11 11 11 11.7 7.67 10.7 41 11.7 8.33 11.3 12 43 3-46

OCR for page 60
Level . Intermediate Schedule Building Chapter 3. Schedule Building A sketch of proposed headway changes (refer back to page ) indicated that an eight-min- ute headway in the peak of the peak would meet the load standard. The tables below show a comparison of present versus proposed adjusted running times. On these tables, the start time of the first trip with a change in running time is highlighted in gray to allow the reader to refer back to the original schedule. NORTHBOUND Current Runtimes Proposed Runtimes Period Period Beginning Wishram Sand Point Beginning Wishram Sand Point : : : : : : : : : : : : : : : : : : : : 3-47

OCR for page 60
Chapter 3. Schedule Building Level . Intermediate Schedule Building SOUTHBOUND Current Runtimes Proposed Runtimes Period Be- Period ginning Wishram Sand Point Beginning Wishram Sand Point : : : : : : : : : : : : : : : : : : : Clearly, the southbound direction running time gets the biggest adjustment. In the PM peak four minutes will be added to both branches over the present allowances. In contrast, only minor changes are made to the : peak running time period in the northbound direction. An additional running time period (from : to : ) is added in the southbound direction, both in order to help ramp up running times smoothly and because added running time is needed that early in the PM. There is an additional running time after the PM peak when running times are dropping by as much as minutes per period. Experience has shown that it is relatively easy to smooth out the running time transition, since the headway will be widening through- out that time as well. In the advanced section, you will take these changes and incorporate them into a fully revised Route schedule, with new running times that differ throughout the day and adjusted head- ways to meet demand. 3-48

OCR for page 60
Level . Intermediate Schedule Building Chapter 3. Schedule Building LEVEL End of Intermediate Schedule Building, 2B Part B. The Advanced Section of Schedule Building continues on the next page. To jump to Schedule Blocking, go to page - . 3-49