1 00:00:00,890 --> 00:00:03,260 The following content is provided under a Creative 2 00:00:03,260 --> 00:00:04,650 Commons License. 3 00:00:04,650 --> 00:00:06,860 Your support will help MIT OpenCourseWare 4 00:00:06,860 --> 00:00:10,950 continue to offer high-quality educational resources for free. 5 00:00:10,950 --> 00:00:13,490 To make a donation or to view additional materials 6 00:00:13,490 --> 00:00:17,450 from hundreds of MIT courses, visit MIT OpenCourseWare 7 00:00:17,450 --> 00:00:18,680 at ocw.mit.edu. 8 00:00:23,070 --> 00:00:27,100 NEEMA NASSIR: So good afternoon, everyone. 9 00:00:27,100 --> 00:00:30,060 Today, we are going to continue the discussion regarding 10 00:00:30,060 --> 00:00:34,570 short-range planning for transit agencies. 11 00:00:34,570 --> 00:00:39,870 And if you do not have the set of slides, 12 00:00:39,870 --> 00:00:43,710 the handout from last week, there are extra copies here. 13 00:00:43,710 --> 00:00:46,800 Please, you can take from here. 14 00:00:46,800 --> 00:00:47,640 OK. 15 00:00:47,640 --> 00:00:49,350 Thanks. 16 00:00:49,350 --> 00:00:50,310 OK. 17 00:00:50,310 --> 00:00:51,720 So last week we-- 18 00:00:51,720 --> 00:00:58,560 last Thursday, we talked about the basic concepts 19 00:00:58,560 --> 00:01:02,310 in the practice of short-range transit planning, 20 00:01:02,310 --> 00:01:07,500 the level of service guidelines and performance measures that 21 00:01:07,500 --> 00:01:09,300 could actually be defined and used 22 00:01:09,300 --> 00:01:14,040 for the practice, particularly related 23 00:01:14,040 --> 00:01:20,460 to excess distances, services span, maximum load, 24 00:01:20,460 --> 00:01:22,140 convenience, and comfort measures. 25 00:01:24,880 --> 00:01:26,460 Today, we will talk a little bit more 26 00:01:26,460 --> 00:01:31,830 about guidelines and standards related to reliability, 27 00:01:31,830 --> 00:01:36,210 and then we will talk about some of the useful performance 28 00:01:36,210 --> 00:01:40,812 measures that could be defined and measured 29 00:01:40,812 --> 00:01:43,020 for transit service. 30 00:01:43,020 --> 00:01:47,880 And then we will talk about the actual procedure 31 00:01:47,880 --> 00:01:55,530 of short-range planning for some of the agencies from which we 32 00:01:55,530 --> 00:01:56,820 have some examples. 33 00:01:56,820 --> 00:02:00,360 And then we will talk about some of the caveats that 34 00:02:00,360 --> 00:02:04,900 could actually be identified for those. 35 00:02:04,900 --> 00:02:11,940 So let's start with guidelines and with the standards related 36 00:02:11,940 --> 00:02:14,650 to reliability first. 37 00:02:14,650 --> 00:02:21,060 So in terms of reliability, as we briefly discussed last week, 38 00:02:21,060 --> 00:02:23,640 most agencies have formal procedures 39 00:02:23,640 --> 00:02:28,455 for monitoring service delivery focusing on time performance. 40 00:02:28,455 --> 00:02:30,420 The on-time performance is usually 41 00:02:30,420 --> 00:02:35,250 considered to be zero minutes early and up to five minutes 42 00:02:35,250 --> 00:02:36,000 late. 43 00:02:36,000 --> 00:02:42,270 As we discussed, early arrivals is not considered on-time. 44 00:02:42,270 --> 00:02:49,320 That could be because early departures basically are not 45 00:02:49,320 --> 00:02:53,040 considered on-time because early departures could actually 46 00:02:53,040 --> 00:02:57,150 have a more negative impact on travel time. 47 00:02:57,150 --> 00:03:01,020 And the other possible reason is that it's probably 48 00:03:01,020 --> 00:03:04,080 early departures are easier to avoid. 49 00:03:04,080 --> 00:03:09,900 When you compare later arrivals or later departures, 50 00:03:09,900 --> 00:03:12,060 possible reasons that could actually 51 00:03:12,060 --> 00:03:14,580 contribute to a late arrival could 52 00:03:14,580 --> 00:03:18,360 be the vehicle getting stuck in traffic, congestion, 53 00:03:18,360 --> 00:03:26,220 or due to high load and high dwell times, which is usually 54 00:03:26,220 --> 00:03:28,350 difficult to avoid. 55 00:03:28,350 --> 00:03:35,070 But when it comes to early arrivals and early departures, 56 00:03:35,070 --> 00:03:41,520 the driver could actually slow down in the upstream segments 57 00:03:41,520 --> 00:03:47,190 to make sure that the arrival is kind of on time. 58 00:03:47,190 --> 00:03:52,380 So about 2/3 of agencies-- there are different thresholds 59 00:03:52,380 --> 00:03:56,010 that are defined in terms of acceptable thresholds 60 00:03:56,010 --> 00:04:03,030 for the percentage of arrivals within this tolerance interval 61 00:04:03,030 --> 00:04:07,140 from zero to five minutes late. 62 00:04:07,140 --> 00:04:12,180 However, the usual practice is that about 2/3 of agencies 63 00:04:12,180 --> 00:04:16,769 report on rush hour, on-time performances of 90% and above. 64 00:04:16,769 --> 00:04:22,740 But in practice, this threshold ranges between 80%, 85%, 65 00:04:22,740 --> 00:04:25,830 up to 90%, 98%, depending on the mode of the service 66 00:04:25,830 --> 00:04:32,100 and depending on the situation, how the congestion level is, 67 00:04:32,100 --> 00:04:37,950 how the load in the network is, if there is right of way, 68 00:04:37,950 --> 00:04:43,800 like separated right of way for that transit service or not. 69 00:04:43,800 --> 00:04:48,150 So it is written somehow as specific to the system, 70 00:04:48,150 --> 00:04:53,580 but 90% is basically the threshold that 2/3 of agencies 71 00:04:53,580 --> 00:04:57,540 basically accept. 72 00:04:57,540 --> 00:05:00,960 Most agencies also keep track and keep route level 73 00:05:00,960 --> 00:05:03,690 information on passenger complaints 74 00:05:03,690 --> 00:05:09,710 that could relate to on-time or late arrivals. 75 00:05:09,710 --> 00:05:13,560 It could relate also to denial to board 76 00:05:13,560 --> 00:05:16,410 or to the crowding situation. 77 00:05:16,410 --> 00:05:22,650 They would also keep track and keep information 78 00:05:22,650 --> 00:05:25,830 on missed trips and accidents and interruptions 79 00:05:25,830 --> 00:05:30,540 and disruptions, basically. 80 00:05:30,540 --> 00:05:34,000 So when it comes to reliability, there 81 00:05:34,000 --> 00:05:37,570 is a need to explain this. 82 00:05:37,570 --> 00:05:39,210 I briefly mentioned this last week, 83 00:05:39,210 --> 00:05:40,910 that the services, the transit service 84 00:05:40,910 --> 00:05:45,250 is basically classified into two types. 85 00:05:45,250 --> 00:05:49,860 One is walk-up service, which is related to higher frequency 86 00:05:49,860 --> 00:05:52,890 service that is usually schedule-free 87 00:05:52,890 --> 00:05:55,800 and/or headway-based. 88 00:05:55,800 --> 00:05:58,650 Typically with a service that is with headway 89 00:05:58,650 --> 00:06:03,570 less than 10 minutes or some say less than 15 minutes, 90 00:06:03,570 --> 00:06:06,060 it's considered to be walk-up service. 91 00:06:06,060 --> 00:06:09,390 For example, the T service in the central area of Boston 92 00:06:09,390 --> 00:06:12,960 would be considered a walk-up service. 93 00:06:12,960 --> 00:06:15,150 BRT systems are also, some of them 94 00:06:15,150 --> 00:06:19,170 are schedule-free and could be considered walk-up 95 00:06:19,170 --> 00:06:22,440 or basically, headway-based. 96 00:06:22,440 --> 00:06:24,270 Performance measures that are related 97 00:06:24,270 --> 00:06:28,410 to reliability for walk-up service is based on headway, 98 00:06:28,410 --> 00:06:32,190 and it basically measures the regularity of headways. 99 00:06:32,190 --> 00:06:35,940 As opposed to service, which is scheduled that's basically 100 00:06:35,940 --> 00:06:41,190 based on published timetables and published schedule 101 00:06:41,190 --> 00:06:45,990 and delivered based on that, basically that's 102 00:06:45,990 --> 00:06:49,300 the service that is usually for longer headway. 103 00:06:49,300 --> 00:06:52,230 Headway is longer than 10 minutes, 10, 15 minutes. 104 00:06:52,230 --> 00:06:55,170 And the performance measure related to this service 105 00:06:55,170 --> 00:06:58,980 is usually being done or being measured 106 00:06:58,980 --> 00:07:02,820 based on punctuality of service, being late or early. 107 00:07:06,790 --> 00:07:10,450 So again, a trip is considered on time 108 00:07:10,450 --> 00:07:13,930 at each time point along the route 109 00:07:13,930 --> 00:07:17,020 and on the performance window, basically. 110 00:07:17,020 --> 00:07:20,260 So usually, transit agencies, for each route 111 00:07:20,260 --> 00:07:22,900 they define certain time points, which 112 00:07:22,900 --> 00:07:27,760 could be major stops or stations but usually, they 113 00:07:27,760 --> 00:07:29,170 are less frequent. 114 00:07:29,170 --> 00:07:31,750 I mean every three or four stops, 115 00:07:31,750 --> 00:07:33,940 there could be one time point that you 116 00:07:33,940 --> 00:07:38,980 measure the on-time arrivals and on-time departures from there. 117 00:07:38,980 --> 00:07:43,330 You may have noticed, if you're on a certain bus, 118 00:07:43,330 --> 00:07:45,100 this is my personal experience, when I 119 00:07:45,100 --> 00:07:47,670 lived in Brisbane in Australia. 120 00:07:47,670 --> 00:07:52,450 There were some routes that at some time when 121 00:07:52,450 --> 00:07:57,970 you see that the bus is early, the driver could actually 122 00:07:57,970 --> 00:08:00,320 really slow down before certain points, 123 00:08:00,320 --> 00:08:03,010 and you could tell where exactly the time point is 124 00:08:03,010 --> 00:08:06,010 that the driver is trying to basically make 125 00:08:06,010 --> 00:08:09,010 an on-time arrival to those time points. 126 00:08:09,010 --> 00:08:13,090 So arrivals are measured at these time 127 00:08:13,090 --> 00:08:15,920 points along the route. 128 00:08:15,920 --> 00:08:16,420 Right. 129 00:08:21,860 --> 00:08:27,440 So what are the actual standards and actual tolerance 130 00:08:27,440 --> 00:08:31,820 or thresholds that agencies consider for service 131 00:08:31,820 --> 00:08:33,590 reliability measures? 132 00:08:33,590 --> 00:08:37,530 One example is here from MBTA. 133 00:08:37,530 --> 00:08:40,549 MBTA, actually for walk-up service, 134 00:08:40,549 --> 00:08:45,700 which is the case for most of the T service, 135 00:08:45,700 --> 00:08:49,450 it allows basically actual headway 136 00:08:49,450 --> 00:08:53,620 of up to 1.5 times the scheduled headway 137 00:08:53,620 --> 00:08:56,200 at origins and at time points. 138 00:08:56,200 --> 00:08:59,560 And then at destinations, it also 139 00:08:59,560 --> 00:09:02,470 defines another measure that deals with the schedule's 140 00:09:02,470 --> 00:09:03,460 running time. 141 00:09:03,460 --> 00:09:05,710 So basically, there are two hooks 142 00:09:05,710 --> 00:09:10,060 that it kind of measures the service based on those. 143 00:09:10,060 --> 00:09:12,550 One is the regularity of service. 144 00:09:12,550 --> 00:09:17,530 It doesn't allow or it wants to measure cases 145 00:09:17,530 --> 00:09:22,420 where headway is longer than expected, like 1.5 times 146 00:09:22,420 --> 00:09:24,010 the scheduled headway. 147 00:09:24,010 --> 00:09:27,640 And the other constraint is that the actual running time 148 00:09:27,640 --> 00:09:31,840 is between 0.8 and 1.2 of the scheduled running time. 149 00:09:34,820 --> 00:09:38,530 So when it comes to actual headway, 150 00:09:38,530 --> 00:09:41,410 you can see that there is no lower bound 151 00:09:41,410 --> 00:09:42,940 to the actual headway. 152 00:09:42,940 --> 00:09:46,030 Does anyone have any comments why there is no actual lower 153 00:09:46,030 --> 00:09:46,660 bound here? 154 00:09:50,540 --> 00:09:53,450 Yeah, please. 155 00:09:53,450 --> 00:09:55,740 AUDIENCE: Lower than scheduled headway just means 156 00:09:55,740 --> 00:09:58,910 that there's more service. 157 00:09:58,910 --> 00:10:01,660 And if that leaves a gap somewhere else, 158 00:10:01,660 --> 00:10:04,560 then that would be record-- that would 159 00:10:04,560 --> 00:10:07,567 count for over the schedule. 160 00:10:07,567 --> 00:10:08,400 NEEMA NASSIR: Right. 161 00:10:08,400 --> 00:10:08,900 Yeah. 162 00:10:08,900 --> 00:10:12,690 Given that the number of trips are constant during the day, 163 00:10:12,690 --> 00:10:18,120 if you guarantee that there is a maximum headway, then 164 00:10:18,120 --> 00:10:19,620 on the other hand, you can kind of 165 00:10:19,620 --> 00:10:25,360 guarantee that headway is not too short there. 166 00:10:25,360 --> 00:10:27,670 However, when it comes to scheduled service, 167 00:10:27,670 --> 00:10:34,290 then there is constraints on early and late departures, 168 00:10:34,290 --> 00:10:35,140 basically. 169 00:10:35,140 --> 00:10:42,430 So at origins, what MBTA considers, at origin terminus, 170 00:10:42,430 --> 00:10:45,700 considers on-time departures between zero and three 171 00:10:45,700 --> 00:10:48,730 minutes, zero minutes early and three minutes late. 172 00:10:48,730 --> 00:10:50,890 And then at time points, it allows up 173 00:10:50,890 --> 00:10:53,620 to seven minutes late. 174 00:10:53,620 --> 00:10:56,920 Does anyone have any comments about the difference? 175 00:10:56,920 --> 00:11:03,190 Why does it allow less late departures 176 00:11:03,190 --> 00:11:04,150 at the origin terminus? 177 00:11:07,348 --> 00:11:11,212 AUDIENCE: Give us control at the intermediary terminus 178 00:11:11,212 --> 00:11:13,160 [INAUDIBLE] at the terminal. 179 00:11:13,160 --> 00:11:14,210 NEEMA NASSIR: Yeah. 180 00:11:14,210 --> 00:11:18,170 That could be the case. 181 00:11:18,170 --> 00:11:23,960 or to put it the other way, transit agencies, 182 00:11:23,960 --> 00:11:27,000 when they schedule these trips and they schedule 183 00:11:27,000 --> 00:11:31,460 these vehicles, they allow a recovery time 184 00:11:31,460 --> 00:11:36,920 between the end of one trip and the beginning of the next trip. 185 00:11:36,920 --> 00:11:39,320 And that actually is there to kind 186 00:11:39,320 --> 00:11:44,390 of help with punctuality of service, 187 00:11:44,390 --> 00:11:51,340 and therefore, it does allow less tolerance on delay 188 00:11:51,340 --> 00:11:52,250 at the beginning. 189 00:11:52,250 --> 00:11:56,780 But when it goes along the route, when 190 00:11:56,780 --> 00:11:59,150 bus goes along the route, it kind of 191 00:11:59,150 --> 00:12:02,860 gets out of control as you mentioned. 192 00:12:02,860 --> 00:12:06,425 There are not really good ways and good leverages 193 00:12:06,425 --> 00:12:12,890 to allow buffer to the running time. 194 00:12:12,890 --> 00:12:16,420 And then at the destination, it's basically, 195 00:12:16,420 --> 00:12:19,330 MBTA allows up to three minutes early 196 00:12:19,330 --> 00:12:23,890 and five minutes late it should be obvious 197 00:12:23,890 --> 00:12:28,840 why it allows early arrivals to destination because it's 198 00:12:28,840 --> 00:12:33,010 the last stop, and we don't need to worry about arriving early 199 00:12:33,010 --> 00:12:35,770 to the final destination. 200 00:12:35,770 --> 00:12:39,580 At the route level, there is one other guideline 201 00:12:39,580 --> 00:12:43,540 that, at the route level, 75% of time points must be on time. 202 00:12:43,540 --> 00:12:49,870 So although for the overall system, 203 00:12:49,870 --> 00:12:54,200 the threshold here was in the range of 90% or 95%, 204 00:12:54,200 --> 00:12:56,770 or between 80%, 85%, and 95%. 205 00:12:56,770 --> 00:13:01,180 But at route level, when you're looking 206 00:13:01,180 --> 00:13:03,850 at the on-time arrivals, it has to be larger 207 00:13:03,850 --> 00:13:08,020 than 75% of the time. 208 00:13:08,020 --> 00:13:11,980 So what will happen if it doesn't-- 209 00:13:11,980 --> 00:13:15,060 a certain route doesn't meet this standard? 210 00:13:19,740 --> 00:13:20,720 Is there any thoughts? 211 00:13:20,720 --> 00:13:22,500 Oh, please. 212 00:13:22,500 --> 00:13:25,420 AUDIENCE: You could shift the schedule for that segment 213 00:13:25,420 --> 00:13:27,330 and republish it and add extra-- 214 00:13:27,330 --> 00:13:28,390 NEEMA NASSIR: Yeah. 215 00:13:28,390 --> 00:13:34,070 So you can increase, basically, the running time 216 00:13:34,070 --> 00:13:37,390 and make sure that you adjust the schedule in a way that 217 00:13:37,390 --> 00:13:41,140 can be delivered punctually. 218 00:13:41,140 --> 00:13:43,910 So that's what's explained in this slide. 219 00:13:43,910 --> 00:13:53,980 So if you observe in some of the routes the performance in terms 220 00:13:53,980 --> 00:13:57,460 of punctuality is poor, then you can take a look 221 00:13:57,460 --> 00:14:00,790 at the historical data and historical performances 222 00:14:00,790 --> 00:14:06,850 in terms of running time, revise the running time 223 00:14:06,850 --> 00:14:11,920 from terminus to terminus and that will actually, 224 00:14:11,920 --> 00:14:16,630 you can basically revise that to be a mean of this distribution, 225 00:14:16,630 --> 00:14:20,020 assuming that the travel times, the running times 226 00:14:20,020 --> 00:14:22,630 are distributed normally. 227 00:14:22,630 --> 00:14:24,272 And go ahead, please. 228 00:14:24,272 --> 00:14:26,230 AUDIENCE: I mean, if it's only a 5% probability 229 00:14:26,230 --> 00:14:29,360 that this will happen but it happens actually 230 00:14:29,360 --> 00:14:30,505 a lot more than that. 231 00:14:30,505 --> 00:14:33,227 It's probably a result of another factor, 232 00:14:33,227 --> 00:14:37,435 like drivers or vehicles or traffic or [INAUDIBLE] 233 00:14:37,435 --> 00:14:38,920 constructions [INAUDIBLE]. 234 00:14:44,044 --> 00:14:45,960 NEEMA NASSIR: Well, that's a good point, yeah. 235 00:14:45,960 --> 00:14:49,410 So there may be, for example, you 236 00:14:49,410 --> 00:14:55,690 know when there is shared right of way with background traffic, 237 00:14:55,690 --> 00:14:58,710 there's a lot of indeterminacy in terms of travel time, 238 00:14:58,710 --> 00:15:01,710 especially if the roadway is congested, 239 00:15:01,710 --> 00:15:07,020 and that could be one possible result for unreliability 240 00:15:07,020 --> 00:15:08,340 of running time. 241 00:15:08,340 --> 00:15:14,940 One possible thing that could be done is like dedicated lanes 242 00:15:14,940 --> 00:15:18,105 to the transit and kind of trying 243 00:15:18,105 --> 00:15:23,450 to improve the performance and match the schedule. 244 00:15:23,450 --> 00:15:25,320 So that could be one other way. 245 00:15:25,320 --> 00:15:29,400 But what I mentioned was that if this is the last thing that we 246 00:15:29,400 --> 00:15:32,940 can do and that we cannot basically improve 247 00:15:32,940 --> 00:15:37,860 the reliability of arrivals, we may actually allow a little bit 248 00:15:37,860 --> 00:15:41,630 of buffer to the travel time. 249 00:15:41,630 --> 00:15:53,090 And that's not only by extending the running time to a job 250 00:15:53,090 --> 00:15:55,120 that we can observe, but also, you 251 00:15:55,120 --> 00:15:58,820 can adjust the recovery time between end 252 00:15:58,820 --> 00:16:03,400 of one trip to the next trip to the chosen amount that 253 00:16:03,400 --> 00:16:09,370 makes sure that a certain confidence can actually 254 00:16:09,370 --> 00:16:13,030 be gained in terms of on-time performance of next trip. 255 00:16:13,030 --> 00:16:18,160 In terms of the actual objective that TransLink 256 00:16:18,160 --> 00:16:21,790 has in place for reliability, again depending 257 00:16:21,790 --> 00:16:26,020 on different modes, TransLink defines 258 00:16:26,020 --> 00:16:34,757 different acceptance thresholds for given on-time definitions. 259 00:16:34,757 --> 00:16:36,340 And on-time definitions could actually 260 00:16:36,340 --> 00:16:43,000 vary as well as a level of compliance or the acceptance 261 00:16:43,000 --> 00:16:45,380 threshold. 262 00:16:45,380 --> 00:16:48,140 So is there any questions about this? 263 00:16:48,140 --> 00:16:48,710 Any comments? 264 00:16:55,370 --> 00:17:04,630 So there are some more detail measures and event 265 00:17:04,630 --> 00:17:11,319 measures for transit service. 266 00:17:11,319 --> 00:17:15,619 It comes to variables that are related to transit service 267 00:17:15,619 --> 00:17:18,400 supply and demand side. 268 00:17:18,400 --> 00:17:22,569 Quantities actually can be classified 269 00:17:22,569 --> 00:17:25,089 into three different groups, the quantities that 270 00:17:25,089 --> 00:17:28,150 are related to the service input, 271 00:17:28,150 --> 00:17:30,100 quantities related to service output, 272 00:17:30,100 --> 00:17:32,950 and quantities related to service consumption. 273 00:17:32,950 --> 00:17:35,890 So service input is basically the resources 274 00:17:35,890 --> 00:17:41,068 that the transit agency is using to produce the service. 275 00:17:41,068 --> 00:17:45,730 Service output is basically measuring the produced service. 276 00:17:45,730 --> 00:17:49,210 And service consumption is how this service 277 00:17:49,210 --> 00:17:52,450 is being consumed and being appreciated and being 278 00:17:52,450 --> 00:17:55,160 used by the public. 279 00:17:55,160 --> 00:18:03,870 So different measures can be defined 280 00:18:03,870 --> 00:18:07,280 to measure the performance of transit system 281 00:18:07,280 --> 00:18:08,880 based on these quantities. 282 00:18:08,880 --> 00:18:13,280 For example, you can look at only the vertices 283 00:18:13,280 --> 00:18:20,420 of this triangle, like what is the amount of labor, including 284 00:18:20,420 --> 00:18:22,490 drivers and maintenance? 285 00:18:22,490 --> 00:18:27,080 What is the capital cost of the fleet and infrastructure? 286 00:18:27,080 --> 00:18:31,850 What is the fuel consumption in the system? 287 00:18:31,850 --> 00:18:36,170 Or, on the other hand, you can look at the service outputs. 288 00:18:36,170 --> 00:18:39,260 It would be like a measure of the service 289 00:18:39,260 --> 00:18:47,220 in terms of how much service the system actually 290 00:18:47,220 --> 00:18:52,100 is producing in terms of vehicle hours or revenue hours, vehicle 291 00:18:52,100 --> 00:18:56,340 miles or revenue hours, capacity miles and service reliability. 292 00:18:56,340 --> 00:18:58,710 And then consumption of service, that 293 00:18:58,710 --> 00:19:00,650 is dealing with number of passengers, 294 00:19:00,650 --> 00:19:05,760 passenger miles, operating revenue, and operating safety. 295 00:19:05,760 --> 00:19:11,940 But you can go one level up with playing with these measures 296 00:19:11,940 --> 00:19:19,650 and compute, for example, the ratios of these quantities. 297 00:19:19,650 --> 00:19:24,200 So each one of these edges basically 298 00:19:24,200 --> 00:19:28,940 are dealing with the ratio of one quantity 299 00:19:28,940 --> 00:19:35,750 at one vertex as opposed to over the quantity of the other one. 300 00:19:35,750 --> 00:19:39,020 For example, what we are basically-- 301 00:19:39,020 --> 00:19:41,180 efficiency measures are basically 302 00:19:41,180 --> 00:19:43,850 dealing with how we are doing in terms 303 00:19:43,850 --> 00:19:49,190 of producing service as given our available resources 304 00:19:49,190 --> 00:19:51,050 or the consumed resources. 305 00:19:51,050 --> 00:19:54,830 The service effectiveness basically 306 00:19:54,830 --> 00:20:00,800 deals with how we are doing in terms of giving service 307 00:20:00,800 --> 00:20:04,400 to public, moving public and producing mobility 308 00:20:04,400 --> 00:20:10,800 in the network over the service that we 309 00:20:10,800 --> 00:20:13,610 produce, how our service is appreciated and used 310 00:20:13,610 --> 00:20:14,870 and utilized. 311 00:20:14,870 --> 00:20:18,800 For example, a service that is poorly designed, for example, 312 00:20:18,800 --> 00:20:23,640 like every five minute trips overnight 313 00:20:23,640 --> 00:20:26,630 from a suburb to another suburb, it's a service, 314 00:20:26,630 --> 00:20:32,780 and it has produced vehicle hours and vehicle miles. 315 00:20:32,780 --> 00:20:34,620 It has consumed resources. 316 00:20:34,620 --> 00:20:36,440 However, it may not be appreciated 317 00:20:36,440 --> 00:20:38,090 because it's poorly designed. 318 00:20:38,090 --> 00:20:42,230 So that's the kind of the role of measures 319 00:20:42,230 --> 00:20:45,050 that are related to service effectiveness 320 00:20:45,050 --> 00:20:51,230 to take care of quality of design and quality of service, 321 00:20:51,230 --> 00:20:53,330 basically. 322 00:20:53,330 --> 00:20:57,500 So and one other comment about this triangle 323 00:20:57,500 --> 00:21:02,210 is that the cost efficiency measures 324 00:21:02,210 --> 00:21:09,800 kind of dealing with the quality of the planning 325 00:21:09,800 --> 00:21:12,875 on the operational side, operational planning. 326 00:21:12,875 --> 00:21:16,560 How the optimization and how the tactical decisions 327 00:21:16,560 --> 00:21:21,650 are made in terms of producing the most service out 328 00:21:21,650 --> 00:21:23,430 of the existing resources. 329 00:21:23,430 --> 00:21:26,450 On the other hand, the decisions that 330 00:21:26,450 --> 00:21:30,530 are related to service planning could actually 331 00:21:30,530 --> 00:21:34,730 use measures that are related to service effectiveness. 332 00:21:34,730 --> 00:21:37,280 So in terms of alternative measures, 333 00:21:37,280 --> 00:21:43,740 like benefit and cost measures that could be defined and used 334 00:21:43,740 --> 00:21:49,110 for short-range transit planning-- 335 00:21:49,110 --> 00:21:52,440 before we go into discussion of benefits and cost 336 00:21:52,440 --> 00:21:56,790 for transit system, there is one thing that I need to mention. 337 00:21:56,790 --> 00:21:59,000 What's happening in transit system 338 00:21:59,000 --> 00:22:03,420 is that they are usually not run, as you know, for profit 339 00:22:03,420 --> 00:22:07,320 and that's probably one of the main differences 340 00:22:07,320 --> 00:22:10,290 between the business model of a transit system as compared 341 00:22:10,290 --> 00:22:13,890 to other regular businesses, where they make decisions 342 00:22:13,890 --> 00:22:20,850 based on profit and based on the revenue over or benefit of cost 343 00:22:20,850 --> 00:22:23,130 or better benefit over cost. 344 00:22:23,130 --> 00:22:26,160 What happens in transit agencies is 345 00:22:26,160 --> 00:22:31,050 that it can be thought as a type of benefit over cost analysis 346 00:22:31,050 --> 00:22:33,970 but the cost and benefit is a little bit different. 347 00:22:33,970 --> 00:22:40,150 The benefit is dealing with the public service and the amount 348 00:22:40,150 --> 00:22:45,730 of consumer surplus or the social welfare 349 00:22:45,730 --> 00:22:49,360 that you can produce, the mobility that you can produce. 350 00:22:49,360 --> 00:22:52,510 On the other hand, the cost could actually 351 00:22:52,510 --> 00:22:57,220 include the operating cost and also could include the revenue. 352 00:22:57,220 --> 00:23:01,850 It basically can be defined as operating cost minus revenue. 353 00:23:01,850 --> 00:23:07,000 So when it comes to the financial unit of benefit 354 00:23:07,000 --> 00:23:11,440 and cost, the financial unit includes both the cost 355 00:23:11,440 --> 00:23:13,190 and revenue. 356 00:23:13,190 --> 00:23:14,780 And then when it comes to benefit, 357 00:23:14,780 --> 00:23:17,840 what's usually being used as a measure of benefit 358 00:23:17,840 --> 00:23:21,170 is the service that is being delivered 359 00:23:21,170 --> 00:23:28,970 in terms of the social welfare measure 360 00:23:28,970 --> 00:23:34,740 or the mobility that actually is produced in the network. 361 00:23:34,740 --> 00:23:39,020 So once we discuss some of these alternative measures, 362 00:23:39,020 --> 00:23:41,935 it may be a little more clear. 363 00:23:41,935 --> 00:23:43,310 So one of the measures that could 364 00:23:43,310 --> 00:23:49,250 be defined and really easy probably to define and measure 365 00:23:49,250 --> 00:23:50,360 is the revenue. 366 00:23:50,360 --> 00:23:52,680 Like, what's the fare box revenue? 367 00:23:52,680 --> 00:23:54,020 What's the total? 368 00:23:54,020 --> 00:23:56,810 The advantage that this measure would have 369 00:23:56,810 --> 00:24:01,400 is that it's very relevant to financial concerns. 370 00:24:01,400 --> 00:24:05,060 It actually, another advantage that it has 371 00:24:05,060 --> 00:24:08,240 is that it's related to willingness to pay 372 00:24:08,240 --> 00:24:13,920 and it could be used as a proxy of quality of service. 373 00:24:13,920 --> 00:24:17,810 However, the disadvantage that could be mentioned for revenue 374 00:24:17,810 --> 00:24:20,390 as a benefit measure is that it discounts 375 00:24:20,390 --> 00:24:24,050 value of reduced fare trips, discounted trips, concession 376 00:24:24,050 --> 00:24:26,420 trips. 377 00:24:26,420 --> 00:24:30,530 And it also has the disadvantage that it favors higher income 378 00:24:30,530 --> 00:24:32,090 passengers. 379 00:24:32,090 --> 00:24:38,550 So if we want to look a little bit more 380 00:24:38,550 --> 00:24:41,220 closely at the nature of transit service 381 00:24:41,220 --> 00:24:44,080 and what should be considered as a benefit measure, 382 00:24:44,080 --> 00:24:47,196 we need to look at the mobility and try 383 00:24:47,196 --> 00:24:52,890 to measure the mobility that has been produced for the citizen, 384 00:24:52,890 --> 00:24:55,510 for the people living in the network. 385 00:24:55,510 --> 00:24:57,900 So we can count the number of passengers 386 00:24:57,900 --> 00:25:02,190 who have used the system and the advantage of this 387 00:25:02,190 --> 00:25:06,330 is that it reflects the number of people who benefit 388 00:25:06,330 --> 00:25:10,490 and it basically values each passenger equally. 389 00:25:10,490 --> 00:25:12,900 However, the shortcoming of counting 390 00:25:12,900 --> 00:25:17,250 only the number of passengers is that it kind of 391 00:25:17,250 --> 00:25:19,290 doesn't reflect trip length. 392 00:25:22,790 --> 00:25:27,240 Not reflecting trip length basically kind of 393 00:25:27,240 --> 00:25:33,000 ignores an important component in mobility, which deals 394 00:25:33,000 --> 00:25:35,730 with access to opportunities. 395 00:25:35,730 --> 00:25:38,910 Which the access to opportunities definitely cannot 396 00:25:38,910 --> 00:25:47,800 decrease with the opportunities for mobility. 397 00:25:47,800 --> 00:25:49,930 And then the other shortcoming is 398 00:25:49,930 --> 00:25:54,340 that it basically reflects on link trips 399 00:25:54,340 --> 00:25:58,950 and not journeys or link trips. 400 00:25:58,950 --> 00:26:01,380 So you basically double count passengers 401 00:26:01,380 --> 00:26:06,270 who are making transfers, although they 402 00:26:06,270 --> 00:26:08,340 may be traveling to one destination 403 00:26:08,340 --> 00:26:10,980 and there's only one opportunity at one destination 404 00:26:10,980 --> 00:26:14,550 or that will be actually one activity happening 405 00:26:14,550 --> 00:26:17,020 at the destination. 406 00:26:17,020 --> 00:26:21,690 So one way to improve this is to measure passenger miles 407 00:26:21,690 --> 00:26:25,200 as a better proxy of mobility in the network. 408 00:26:25,200 --> 00:26:28,800 It has the advantage that it weights longer trips more 409 00:26:28,800 --> 00:26:37,080 and it's the best reflection of benefits related to mobility. 410 00:26:37,080 --> 00:26:38,680 Sorry about this. 411 00:26:38,680 --> 00:26:41,910 However, it's hardest to measure. 412 00:26:41,910 --> 00:26:47,900 Does anyone have any idea, comments about why it's hardest 413 00:26:47,900 --> 00:26:49,900 to measure? 414 00:26:49,900 --> 00:26:56,330 Why is passenger miles in the network hardest to measure? 415 00:26:56,330 --> 00:26:59,115 AUDIENCE: A lot of systems don't have an exit. 416 00:26:59,115 --> 00:27:00,580 They only have an entrance stile. 417 00:27:00,580 --> 00:27:01,750 NEEMA NASSIR: Exactly. 418 00:27:01,750 --> 00:27:06,760 So considering that we're using the data for passenger, 419 00:27:06,760 --> 00:27:09,310 like from fare card systems, many systems 420 00:27:09,310 --> 00:27:13,660 do not have the requirements of tapping 421 00:27:13,660 --> 00:27:15,430 when you're exiting the system. 422 00:27:15,430 --> 00:27:17,390 So that's something that is missing. 423 00:27:17,390 --> 00:27:20,560 So it requires some sort of inference 424 00:27:20,560 --> 00:27:25,310 of highlighting our exit locations to be able to measure 425 00:27:25,310 --> 00:27:26,680 the length of trip. 426 00:27:26,680 --> 00:27:29,590 However, and the other disadvantage 427 00:27:29,590 --> 00:27:34,180 is that it could find favor higher income passengers. 428 00:27:34,180 --> 00:27:36,340 So any comments about this? 429 00:27:36,340 --> 00:27:39,320 AUDIENCE: That assumes the high-income passengers, 430 00:27:39,320 --> 00:27:42,803 that assumes higher income people travel further. 431 00:27:42,803 --> 00:27:45,257 It's not necessarily the case. 432 00:27:45,257 --> 00:27:46,840 NEEMA NASSIR: Not necessarily the case 433 00:27:46,840 --> 00:27:54,820 but in systems where you have distance-based fare structure, 434 00:27:54,820 --> 00:28:00,580 and when you basically have the entrance and exits recorded 435 00:28:00,580 --> 00:28:06,010 and you charge people by the length or number of zones 436 00:28:06,010 --> 00:28:09,700 that they have traveled, that could be actually 437 00:28:09,700 --> 00:28:14,350 somehow favoring higher income passengers 438 00:28:14,350 --> 00:28:17,650 because it is reasonable to assume that higher income 439 00:28:17,650 --> 00:28:21,520 people are less sensitive to the fare 440 00:28:21,520 --> 00:28:26,110 and so their decisions are basically 441 00:28:26,110 --> 00:28:28,470 kind of regardless of fare. 442 00:28:28,470 --> 00:28:30,080 So any comments? 443 00:28:35,554 --> 00:28:36,054 OK. 444 00:28:39,470 --> 00:28:44,170 So then on the cost side, the measures that 445 00:28:44,170 --> 00:28:49,430 are dealing with cost, the net cost 446 00:28:49,430 --> 00:28:53,900 is basically cost of resources that 447 00:28:53,900 --> 00:29:00,770 are used to generate the service minus the revenue. 448 00:29:00,770 --> 00:29:07,970 And it's usually most directly constrained 449 00:29:07,970 --> 00:29:12,330 because that's exactly the amount of money 450 00:29:12,330 --> 00:29:16,680 that you need to secure for the service. 451 00:29:16,680 --> 00:29:21,360 The budget actually goes directly to the net cost. 452 00:29:21,360 --> 00:29:23,980 So that's the advantage. 453 00:29:23,980 --> 00:29:27,373 However, the disadvantage is that it's hardest to estimate. 454 00:29:30,710 --> 00:29:34,800 Any comments about why it could be hardest to estimate as 455 00:29:34,800 --> 00:29:35,625 compared to cost? 456 00:29:41,630 --> 00:29:47,930 At least for designs, for the service that is not already 457 00:29:47,930 --> 00:29:54,500 in place, if you need to estimate the revenue 458 00:29:54,500 --> 00:29:59,420 and deduct that from cost to generate net cost, 459 00:29:59,420 --> 00:30:02,030 it would be a challenging thing to do, 460 00:30:02,030 --> 00:30:06,890 to come up with estimates of demand and predict the revenue. 461 00:30:06,890 --> 00:30:09,080 So that could be one of the shortcomings 462 00:30:09,080 --> 00:30:12,290 of net cost for designs. 463 00:30:12,290 --> 00:30:14,480 However, it may not be that difficult when 464 00:30:14,480 --> 00:30:17,690 we are dealing with the existing service 465 00:30:17,690 --> 00:30:21,350 because revenue is basically already recorded 466 00:30:21,350 --> 00:30:24,610 by the fare box system. 467 00:30:24,610 --> 00:30:29,210 And then cost could be another measure less directly 468 00:30:29,210 --> 00:30:36,170 constrained because actually, we need to combine that 469 00:30:36,170 --> 00:30:37,910 with the revenue to realize what's 470 00:30:37,910 --> 00:30:40,070 the net cost of the system? 471 00:30:40,070 --> 00:30:43,160 Again, it could be hard to estimate 472 00:30:43,160 --> 00:30:48,260 but probably not as hard when we compare it with net cost. 473 00:30:48,260 --> 00:30:49,118 Go ahead. 474 00:30:49,118 --> 00:30:51,410 AUDIENCE: What is net cost? 475 00:30:51,410 --> 00:30:56,540 NEEMA NASSIR: Net cost is basically cost of operation 476 00:30:56,540 --> 00:31:00,710 minus the revenue from fares. 477 00:31:00,710 --> 00:31:04,730 And the reason that it's tricky to measure 478 00:31:04,730 --> 00:31:09,833 is that the revenue side is probably a little difficult 479 00:31:09,833 --> 00:31:10,815 to predict. 480 00:31:15,240 --> 00:31:21,630 So the next measure of cost could be vehicle miles. 481 00:31:21,630 --> 00:31:25,260 That basically deals with the service that is produced. 482 00:31:25,260 --> 00:31:28,890 It's easy to measure because you know number 483 00:31:28,890 --> 00:31:33,870 of trips that are delivered to the system, and length of trips 484 00:31:33,870 --> 00:31:34,380 are known. 485 00:31:34,380 --> 00:31:42,210 So vehicle miles is probably the easiest measure to compute. 486 00:31:42,210 --> 00:31:44,490 But the disadvantage is that they only 487 00:31:44,490 --> 00:31:49,890 reflect 3% of the cost and that's probably 488 00:31:49,890 --> 00:31:55,095 the part that is correlated with fuel and with depreciation 489 00:31:55,095 --> 00:31:59,420 of the vehicles and with maintenance. 490 00:31:59,420 --> 00:32:02,680 And then it also penalizes fast service, 491 00:32:02,680 --> 00:32:09,350 not [AUDIO OUT] traveled fast and on time 492 00:32:09,350 --> 00:32:11,270 as opposed to the model that is basically 493 00:32:11,270 --> 00:32:16,940 delivered in congestion and with long delays. 494 00:32:16,940 --> 00:32:21,740 Another measure that is probably a little bit more 495 00:32:21,740 --> 00:32:26,120 closely reflecting our cost is the vehicle hours. 496 00:32:26,120 --> 00:32:28,370 And the vehicle miles and vehicle hours 497 00:32:28,370 --> 00:32:33,290 are basically the revenue miles and the revenue hours. 498 00:32:33,290 --> 00:32:37,106 What I mean by revenue, that's basically the miles 499 00:32:37,106 --> 00:32:40,190 that these buses and these vehicles 500 00:32:40,190 --> 00:32:44,960 have been in service from the start of their trip 501 00:32:44,960 --> 00:32:46,490 to the end of the trip. 502 00:32:46,490 --> 00:32:54,960 There may be some trips, like dead-ending trips, 503 00:32:54,960 --> 00:32:58,730 that the vehicle can should go empty from one 504 00:32:58,730 --> 00:33:02,930 location to another location to the next trip, 505 00:33:02,930 --> 00:33:05,120 and those miles and those vehicle hours 506 00:33:05,120 --> 00:33:07,340 are not counted here. 507 00:33:07,340 --> 00:33:11,180 So vehicle hours basically, again, are easy to measure. 508 00:33:11,180 --> 00:33:17,180 And the other advantage is that it's related to more than 50% 509 00:33:17,180 --> 00:33:20,870 of bus costs. 510 00:33:20,870 --> 00:33:27,440 It basically correlates with the fuel consumption 511 00:33:27,440 --> 00:33:32,720 and with driving and operating cost. 512 00:33:36,120 --> 00:33:39,530 So one of the cost effectiveness measures 513 00:33:39,530 --> 00:33:45,950 that MBTA uses for their system and for their decision making 514 00:33:45,950 --> 00:33:51,050 is basically net cost per each passenger that 515 00:33:51,050 --> 00:33:53,690 is using the system. 516 00:33:53,690 --> 00:33:59,060 And then they usually define this on a route basis. 517 00:33:59,060 --> 00:34:05,060 So the index I here could be net cost per passenger 518 00:34:05,060 --> 00:34:10,449 for each route, and then it shouldn't exceed three times 519 00:34:10,449 --> 00:34:16,350 the system-wide net cost per passenger. 520 00:34:16,350 --> 00:34:21,739 So this is basically going back to the triangle that we had. 521 00:34:21,739 --> 00:34:25,010 This is one of the measures that is dealing with resources that 522 00:34:25,010 --> 00:34:27,679 are used to the number of passengers 523 00:34:27,679 --> 00:34:30,850 that are using the service, regardless 524 00:34:30,850 --> 00:34:32,080 of what the service was. 525 00:34:36,630 --> 00:34:41,310 So in terms of issues in setting up a short-range transit 526 00:34:41,310 --> 00:34:49,800 planning process, this slide mentions some of these issues, 527 00:34:49,800 --> 00:34:51,540 starting with the roll-off budget 528 00:34:51,540 --> 00:34:54,870 constraints in the process this could actually 529 00:34:54,870 --> 00:34:58,410 be related to before budget is set and after budget is set. 530 00:34:58,410 --> 00:35:02,400 Before budget is set, there is some indeterminacy 531 00:35:02,400 --> 00:35:08,520 in measuring the net cost and basically 532 00:35:08,520 --> 00:35:12,480 the budget that is requested in the proposal for a new service 533 00:35:12,480 --> 00:35:15,570 or for a change in the service. 534 00:35:15,570 --> 00:35:21,300 What's similar to any other proposal that is usually done 535 00:35:21,300 --> 00:35:25,710 is that you try to find the minimum possible budget 536 00:35:25,710 --> 00:35:30,740 that you're requesting for what you're trying to deliver 537 00:35:30,740 --> 00:35:33,880 and that could be a challenging task. 538 00:35:33,880 --> 00:35:35,790 And then after the budget is set, 539 00:35:35,790 --> 00:35:39,960 given all the indeterminacy that was involved in predicting 540 00:35:39,960 --> 00:35:47,850 the budget, then it's the issue of optimizing the resources 541 00:35:47,850 --> 00:35:51,870 and optimizing the budget to deliver and to gain the most 542 00:35:51,870 --> 00:35:57,270 out of the existing resources. 543 00:35:57,270 --> 00:35:59,810 The other issue with-- 544 00:35:59,810 --> 00:36:02,810 AUDIENCE: So in this part of the process, 545 00:36:02,810 --> 00:36:06,970 I wonder if it should be advisable to leave 546 00:36:06,970 --> 00:36:12,450 a margin of maneuver in respect to budget constraints normally 547 00:36:12,450 --> 00:36:16,890 you would have, even if you go for the minimum possible to get 548 00:36:16,890 --> 00:36:21,740 85%, 80%, out of that, for political reasons 549 00:36:21,740 --> 00:36:24,870 or whatever reason there is. 550 00:36:24,870 --> 00:36:26,840 NEEMA NASSIR: Right. 551 00:36:26,840 --> 00:36:32,065 So that could actually make the situation a little easier 552 00:36:32,065 --> 00:36:34,430 for people who are involved in planning 553 00:36:34,430 --> 00:36:36,440 and for the actual plans, right? 554 00:36:36,440 --> 00:36:43,880 So what Ricardo is mentioning is there may be 555 00:36:43,880 --> 00:36:50,570 some buffers or some internal-- 556 00:36:50,570 --> 00:36:56,150 some buffers in the budget and some flexibility 557 00:36:56,150 --> 00:36:58,310 in the budget that could actually 558 00:36:58,310 --> 00:37:02,550 be allowed when you're requesting the budget 559 00:37:02,550 --> 00:37:04,850 and when you're delivering the budget, 560 00:37:04,850 --> 00:37:08,390 delivering the service that could actually 561 00:37:08,390 --> 00:37:10,700 take care of some of the indeterminacy that 562 00:37:10,700 --> 00:37:12,740 is involved in the prediction of the budget 563 00:37:12,740 --> 00:37:13,970 and prediction of the costs. 564 00:37:18,090 --> 00:37:26,500 So the next, possibly the next issue 565 00:37:26,500 --> 00:37:31,600 that is involved in the traditional or the existing 566 00:37:31,600 --> 00:37:34,630 practice for short-range transit planning 567 00:37:34,630 --> 00:37:41,290 is that we kind of have a set of standards and constraints 568 00:37:41,290 --> 00:37:45,780 that gives us lower bounds on some of the service measures. 569 00:37:45,780 --> 00:37:48,900 However, these constraints only give us 570 00:37:48,900 --> 00:37:52,320 an idea of what's the feasible solution space. 571 00:37:52,320 --> 00:37:56,370 It doesn't really tell us what's the optimal design 572 00:37:56,370 --> 00:38:00,648 and how you can basically optimize 573 00:38:00,648 --> 00:38:05,445 the allocation of resources to improve the ridership. 574 00:38:05,445 --> 00:38:09,810 And that could actually involve some 575 00:38:09,810 --> 00:38:12,570 of the art of the designers and the art of people 576 00:38:12,570 --> 00:38:15,870 who are involved in identifying opportunities 577 00:38:15,870 --> 00:38:20,520 within the standards and within the constraints and guidelines 578 00:38:20,520 --> 00:38:23,400 to address the situation and to gain the most 579 00:38:23,400 --> 00:38:28,420 out of the allocated resources. 580 00:38:28,420 --> 00:38:33,880 Then again, when it comes to design of service, 581 00:38:33,880 --> 00:38:38,350 there is also this issue between the existing 582 00:38:38,350 --> 00:38:44,515 service and proposals and plans for additional service. 583 00:38:47,020 --> 00:38:52,030 As I mentioned last week, the design of service 584 00:38:52,030 --> 00:38:53,640 is never from scratch. 585 00:38:53,640 --> 00:38:59,500 There is always something there for good reasons 586 00:38:59,500 --> 00:39:03,490 and you may want to predict the existing service 587 00:39:03,490 --> 00:39:06,160 as much as possible because people 588 00:39:06,160 --> 00:39:07,480 are relying on the service. 589 00:39:07,480 --> 00:39:15,310 They probably have formed their travel habits and travel 590 00:39:15,310 --> 00:39:17,740 attitudes based on the existing service. 591 00:39:17,740 --> 00:39:21,330 So changing these services probably 592 00:39:21,330 --> 00:39:29,080 need to be really careful and thoughtful, basically. 593 00:39:29,080 --> 00:39:35,170 However, when it comes to addition of new service, 594 00:39:35,170 --> 00:39:40,330 we may want to make sure that the new service 595 00:39:40,330 --> 00:39:46,150 kind of supports the existing service effectively 596 00:39:46,150 --> 00:39:51,760 and with all the overlaps and all the misconnections 597 00:39:51,760 --> 00:39:55,120 are identified in advance and taken care of. 598 00:39:58,390 --> 00:40:01,240 Issue that we will eventually talk about it later 599 00:40:01,240 --> 00:40:06,580 in another slide is that some of existing practice 600 00:40:06,580 --> 00:40:11,950 is dealing with short-range transit planning decisions 601 00:40:11,950 --> 00:40:16,720 dealing poorly performing routes. 602 00:40:16,720 --> 00:40:19,750 But the optimal design is not always 603 00:40:19,750 --> 00:40:23,570 dealing with changing the ones that are only poor, 604 00:40:23,570 --> 00:40:27,070 the only ones that are poorly performing. 605 00:40:27,070 --> 00:40:29,320 There may be some changes that you 606 00:40:29,320 --> 00:40:35,980 can do to well-performing routes to improve the overall system 607 00:40:35,980 --> 00:40:39,340 performance and the network level performance. 608 00:40:41,870 --> 00:40:43,730 So is there any questions so far? 609 00:40:47,980 --> 00:40:49,134 How am I doing on time? 610 00:40:49,134 --> 00:40:50,902 OK. 611 00:40:50,902 --> 00:40:53,920 AUDIENCE: Professor Nassir, does the cost effectiveness standard 612 00:40:53,920 --> 00:40:58,537 work if they can't compare the cost 613 00:40:58,537 --> 00:41:01,680 effectiveness of [INAUDIBLE]? 614 00:41:01,680 --> 00:41:03,090 NEEMA NASSIR: My sense is that it 615 00:41:03,090 --> 00:41:12,570 has to be in a modus specific, but yeah, 616 00:41:12,570 --> 00:41:15,770 but I can take that and-- 617 00:41:15,770 --> 00:41:16,340 Yeah. 618 00:41:16,340 --> 00:41:21,980 It makes more sense to be, yeah, because you probably cannot 619 00:41:21,980 --> 00:41:28,100 compare the effectiveness of a bus versus a train in terms 620 00:41:28,100 --> 00:41:29,910 of operating costs at least, right? 621 00:41:35,790 --> 00:41:43,730 So this is the standard process of decision 622 00:41:43,730 --> 00:41:46,820 making when it comes to new projects and new proposals, 623 00:41:46,820 --> 00:41:53,822 service proposals, and changes to the existing service 624 00:41:53,822 --> 00:41:57,170 in Toronto Transit Commission. 625 00:41:57,170 --> 00:42:00,740 This gives you an overview of different tasks 626 00:42:00,740 --> 00:42:06,170 that are basically done to start with a set of proposals 627 00:42:06,170 --> 00:42:12,690 and to make the decision on which ones to be implemented 628 00:42:12,690 --> 00:42:14,190 and so on and so forth. 629 00:42:14,190 --> 00:42:18,250 So if you look at the column on the right, 630 00:42:18,250 --> 00:42:21,650 it basically includes the decision-- 631 00:42:21,650 --> 00:42:23,860 it deals mostly with the decision making 632 00:42:23,860 --> 00:42:28,400 on the new service proposals. 633 00:42:28,400 --> 00:42:35,080 First, you get some requests from municipalities 634 00:42:35,080 --> 00:42:38,920 and some suggestions from the staff 635 00:42:38,920 --> 00:42:42,370 about new service proposals. 636 00:42:42,370 --> 00:42:50,470 Then, if these proposals are passing your system guidelines, 637 00:42:50,470 --> 00:42:53,030 then there would be two possibilities. 638 00:42:53,030 --> 00:42:57,290 If they are minor costs, they'd go directly to recommendation. 639 00:42:57,290 --> 00:42:59,350 And if they are major costs, then you 640 00:42:59,350 --> 00:43:01,750 need to do a benefit cost analysis, 641 00:43:01,750 --> 00:43:04,600 and that basically includes estimation 642 00:43:04,600 --> 00:43:09,400 of ridership and estimation of costs and benefits. 643 00:43:09,400 --> 00:43:15,150 And then, once you have a better sense of benefits and costs 644 00:43:15,150 --> 00:43:20,840 and benefit/cost measures of the project, 645 00:43:20,840 --> 00:43:24,010 then there will be a comparative evaluation 646 00:43:24,010 --> 00:43:28,570 of alternative designs and alternative proposals. 647 00:43:28,570 --> 00:43:31,060 At this stage, there will be another input 648 00:43:31,060 --> 00:43:35,170 from changes to the existing system that 649 00:43:35,170 --> 00:43:39,010 comes into the comparative evaluation, which we will 650 00:43:39,010 --> 00:43:42,840 get to that branch in a minute. 651 00:43:42,840 --> 00:43:48,610 After these projects get to the recommendation level, 652 00:43:48,610 --> 00:43:52,590 the staff makes a recommendation to the board, 653 00:43:52,590 --> 00:43:57,360 and then there will be a review and decision 654 00:43:57,360 --> 00:44:02,160 about acceptance or rejections of some of these proposals. 655 00:44:02,160 --> 00:44:04,680 And then there will be an implementation 656 00:44:04,680 --> 00:44:07,140 of the successful proposals. 657 00:44:07,140 --> 00:44:10,842 And then there will be a probationary time 658 00:44:10,842 --> 00:44:16,110 to implement the successful proposals and then to, 659 00:44:16,110 --> 00:44:21,240 after six months, it will get back to the recommender team 660 00:44:21,240 --> 00:44:27,510 to see if it would pass the probationary period. 661 00:44:27,510 --> 00:44:31,790 On the other hand, there is a set 662 00:44:31,790 --> 00:44:35,420 of proposals that are made to make changes to the existing 663 00:44:35,420 --> 00:44:39,990 service and that basically comes from two main streams-- 664 00:44:39,990 --> 00:44:45,290 one is a continuous monitoring of ridership and loads 665 00:44:45,290 --> 00:44:49,700 and the other one is the annual route efficiency review. 666 00:44:49,700 --> 00:44:54,350 And from this, then the staff actually 667 00:44:54,350 --> 00:44:58,040 identify service changes and assess the impacts 668 00:44:58,040 --> 00:44:59,930 on passengers. 669 00:44:59,930 --> 00:45:03,260 If the impacts are minor, it goes directly 670 00:45:03,260 --> 00:45:04,940 and it becomes implemented. 671 00:45:04,940 --> 00:45:08,560 If impacts are major, then these changes 672 00:45:08,560 --> 00:45:12,110 become into a form of a proposal and comes 673 00:45:12,110 --> 00:45:16,130 to the stream of comparative analysis 674 00:45:16,130 --> 00:45:21,200 that would be actually compared with new service proposals 675 00:45:21,200 --> 00:45:26,071 and follow the same procedure with that. 676 00:45:26,071 --> 00:45:26,570 OK. 677 00:45:26,570 --> 00:45:27,611 Any questions about this? 678 00:45:34,780 --> 00:45:37,150 We'll actually get to the main two streams 679 00:45:37,150 --> 00:45:42,610 on changes to the existing service 680 00:45:42,610 --> 00:45:45,770 in the second next slide right now. 681 00:45:45,770 --> 00:45:49,510 So the first slide is dealing with basically 682 00:45:49,510 --> 00:45:53,110 the continuous monitoring of the system 683 00:45:53,110 --> 00:45:58,330 that Transit Authority or transit agency staff 684 00:45:58,330 --> 00:45:59,980 regularly do. 685 00:45:59,980 --> 00:46:05,170 It's basically, the input into this kind of process 686 00:46:05,170 --> 00:46:09,700 involves ridership counts, regular ridership counts, 687 00:46:09,700 --> 00:46:13,750 customer communication, customer complaint records, 688 00:46:13,750 --> 00:46:16,930 some observations from staff in terms 689 00:46:16,930 --> 00:46:18,920 of the performance of these routes 690 00:46:18,920 --> 00:46:24,250 and for possible adjustments that 691 00:46:24,250 --> 00:46:27,790 could be done and then from operations report. 692 00:46:27,790 --> 00:46:35,290 Then there will be a comparison of ridership to load standards. 693 00:46:35,290 --> 00:46:37,600 What we basically discussed last week 694 00:46:37,600 --> 00:46:41,290 in terms of level of service measures related 695 00:46:41,290 --> 00:46:48,220 to maximum load will actually be tested for all the routes. 696 00:46:48,220 --> 00:46:54,080 And then depending on the situation, 697 00:46:54,080 --> 00:47:02,320 if there is available vehicles and budgets, then the staff 698 00:47:02,320 --> 00:47:06,390 actually making their recommendations of changing 699 00:47:06,390 --> 00:47:09,280 the service, if there is availability of vehicle 700 00:47:09,280 --> 00:47:12,120 and budget, there will be improvement in the service 701 00:47:12,120 --> 00:47:14,280 and increase in the service level. 702 00:47:14,280 --> 00:47:18,520 And then it will be reviewed for recommendation, staff approval, 703 00:47:18,520 --> 00:47:23,230 and implementation, and this is a closed loop that is usually 704 00:47:23,230 --> 00:47:26,210 done on a regular basis. 705 00:47:26,210 --> 00:47:28,720 Then there is another process related 706 00:47:28,720 --> 00:47:30,490 to the existing service that deals 707 00:47:30,490 --> 00:47:38,290 with yearly review of the routes in terms of route efficiency 708 00:47:38,290 --> 00:47:39,430 review. 709 00:47:39,430 --> 00:47:45,070 And basically, you kind of consider a route 710 00:47:45,070 --> 00:47:47,990 from a cost and benefit perspective. 711 00:47:47,990 --> 00:47:53,740 And on the left, you can see new riding counts or yearly riding 712 00:47:53,740 --> 00:47:59,110 counts that's actually are collected from the study year. 713 00:47:59,110 --> 00:48:05,170 And then that goes complemented with the record 714 00:48:05,170 --> 00:48:07,720 of complaints and customer satisfaction 715 00:48:07,720 --> 00:48:10,690 reports and service regularity. 716 00:48:10,690 --> 00:48:15,520 And it will be combined with the financial performance 717 00:48:15,520 --> 00:48:17,710 and basically the measure of costs 718 00:48:17,710 --> 00:48:20,230 that are spent to each route. 719 00:48:20,230 --> 00:48:28,340 And then it comes to this joint here with reviewing 720 00:48:28,340 --> 00:48:29,960 the structure of the route. 721 00:48:29,960 --> 00:48:32,720 Some of these routes may be routes 722 00:48:32,720 --> 00:48:37,250 with branches which actually makes the analysis a little bit 723 00:48:37,250 --> 00:48:41,500 more difficult, especially design 724 00:48:41,500 --> 00:48:46,250 of branching and alignment of branching, making sure 725 00:48:46,250 --> 00:48:56,070 that the frequency and the load profiles and maximum loads 726 00:48:56,070 --> 00:48:59,460 are actually satisfying the standards 727 00:48:59,460 --> 00:49:00,840 before and after the joints. 728 00:49:00,840 --> 00:49:05,490 Because as you may have already thought about, 729 00:49:05,490 --> 00:49:11,070 some of-- so, for example, say Green Line here in Boston, 730 00:49:11,070 --> 00:49:12,450 in the part that is-- 731 00:49:19,060 --> 00:49:26,760 for example, where you have higher demand 732 00:49:26,760 --> 00:49:31,780 in the central area, one of the effective techniques 733 00:49:31,780 --> 00:49:38,470 in design of routes is to join some of these routes 734 00:49:38,470 --> 00:49:42,370 into one service to avoid additional transfers 735 00:49:42,370 --> 00:49:48,760 and to produce connectivity to the central area 736 00:49:48,760 --> 00:49:51,250 without need of a transfer here. 737 00:49:51,250 --> 00:49:55,240 However what it actually causes is 738 00:49:55,240 --> 00:50:01,840 that the demand that you're collecting from lower density 739 00:50:01,840 --> 00:50:11,240 suburbs are basically based on given frequencies here. 740 00:50:11,240 --> 00:50:14,330 And you set these frequency based on these demands. 741 00:50:14,330 --> 00:50:17,750 But when it comes to the merge points, 742 00:50:17,750 --> 00:50:20,480 the frequency will be the combined frequency 743 00:50:20,480 --> 00:50:22,850 and the demand, kind of crowded situation 744 00:50:22,850 --> 00:50:24,590 will be the combined situation. 745 00:50:24,590 --> 00:50:28,250 So you need to make sure that all these standards and all 746 00:50:28,250 --> 00:50:35,802 the guidelines are holding for all these branches 747 00:50:35,802 --> 00:50:36,510 at the same time. 748 00:50:39,800 --> 00:50:41,210 OK. 749 00:50:41,210 --> 00:50:48,360 And then after that, let me try to clarify this a little bit. 750 00:50:48,360 --> 00:50:54,090 For example, considering the-- 751 00:50:54,090 --> 00:50:56,190 similar to the example that we made 752 00:50:56,190 --> 00:51:04,320 last week related to two suburbs that are actually 753 00:51:04,320 --> 00:51:07,620 using the transit service. 754 00:51:11,550 --> 00:51:20,670 So let's say suburb one, suburb two, and this is downtown area. 755 00:51:20,670 --> 00:51:24,690 If the existing service is a direct line between suburb one 756 00:51:24,690 --> 00:51:30,840 to destination and passengers from suburb two 757 00:51:30,840 --> 00:51:35,730 basically walk to this stop to get to destination, 758 00:51:35,730 --> 00:51:43,680 then assume that the new proposal of service 759 00:51:43,680 --> 00:51:56,440 is basically adding one detour here to serve this. 760 00:51:56,440 --> 00:51:58,730 Try to clarify this a little bit. 761 00:51:58,730 --> 00:52:05,540 For example, considering similar to the example 762 00:52:05,540 --> 00:52:13,990 that we made last week related to two suburbs that 763 00:52:13,990 --> 00:52:21,920 are actually using the transit service in the-- 764 00:52:21,920 --> 00:52:31,050 so let's say suburb one, suburb two, and this is downtown area. 765 00:52:31,050 --> 00:52:35,040 If the existing service is a direct line between suburb one 766 00:52:35,040 --> 00:52:41,190 to destination and passengers from suburb two 767 00:52:41,190 --> 00:52:46,110 basically walk to this stop to get to destination, 768 00:52:46,110 --> 00:52:54,060 then assume that the new proposal of service 769 00:52:54,060 --> 00:53:06,690 is basically adding one detour here to serve this suburb 770 00:53:06,690 --> 00:53:11,700 two on the way to destination, then what needs to be done 771 00:53:11,700 --> 00:53:16,620 is that we need to make sure that the total generalized cost 772 00:53:16,620 --> 00:53:19,980 that is computed for all the passengers from suburb 773 00:53:19,980 --> 00:53:23,460 one and suburb two to destination 774 00:53:23,460 --> 00:53:31,080 would be reduced after the introduction of this detour. 775 00:53:31,080 --> 00:53:36,540 And then how do we compute the total generalized cost 776 00:53:36,540 --> 00:53:39,030 that is experienced by the passengers? 777 00:53:39,030 --> 00:53:44,320 There are some equivalent values that are usually used. 778 00:53:44,320 --> 00:53:47,460 For example, for every vehicle travel time, 779 00:53:47,460 --> 00:53:52,230 you assume, one in vehicle travel minute. 780 00:53:52,230 --> 00:53:55,110 And then for every minute of waiting time, 781 00:53:55,110 --> 00:53:57,900 you assume 1.5 or so. 782 00:53:57,900 --> 00:54:01,440 For every minute of walking time that is required, 783 00:54:01,440 --> 00:54:04,170 you consider two minutes of in-vehicle time. 784 00:54:04,170 --> 00:54:06,130 That's the case for, for example, 785 00:54:06,130 --> 00:54:12,006 when you want to measure the cost of walking from this area 786 00:54:12,006 --> 00:54:14,300 to the stop. 787 00:54:14,300 --> 00:54:16,550 And then the cost of each transfer 788 00:54:16,550 --> 00:54:24,790 is considered to be 10 minutes, equivalent to 10 minutes. 789 00:54:24,790 --> 00:54:27,630 Is there any questions about this? 790 00:54:27,630 --> 00:54:34,402 AUDIENCE: So from the local suburb [INAUDIBLE] so in case 791 00:54:34,402 --> 00:54:38,938 there would be [INAUDIBLE] However, 792 00:54:38,938 --> 00:54:43,660 it may just be the fact that [INAUDIBLE].. 793 00:54:43,660 --> 00:54:45,460 NEEMA NASSIR: Right. 794 00:54:45,460 --> 00:54:48,260 If the service can attract latent demand-- 795 00:54:48,260 --> 00:54:54,380 so the question was that this condition does not 796 00:54:54,380 --> 00:54:56,690 account for latent demand, and it kind of 797 00:54:56,690 --> 00:55:00,710 assumes that the demand is constant between the two 798 00:55:00,710 --> 00:55:02,540 scenarios. 799 00:55:02,540 --> 00:55:05,330 So my impression is that whenever 800 00:55:05,330 --> 00:55:07,460 there is latent demand, that could actually 801 00:55:07,460 --> 00:55:09,342 be considered an advantage. 802 00:55:09,342 --> 00:55:19,120 So you can safely assume with the existing level of demand 803 00:55:19,120 --> 00:55:24,860 if this change or if this new service can 804 00:55:24,860 --> 00:55:28,014 reduce the total system cost? 805 00:55:28,014 --> 00:55:32,590 But I do not know how they exactly 806 00:55:32,590 --> 00:55:37,066 implement this when it comes to demand prediction. 807 00:55:41,034 --> 00:55:43,030 OK. 808 00:55:43,030 --> 00:55:45,230 So any other questions about this? 809 00:55:45,230 --> 00:55:48,910 You may ask where does these numbers come from? 810 00:55:48,910 --> 00:55:54,280 I mean, these coefficients, how do we get these coefficients? 811 00:55:54,280 --> 00:55:55,400 They may make sense. 812 00:55:55,400 --> 00:55:57,100 At least, to me, actually, they're 813 00:55:57,100 --> 00:56:01,660 kind of consistent with the way I perceive the cost of travel 814 00:56:01,660 --> 00:56:06,820 in different components of trips, in vehicle time, 815 00:56:06,820 --> 00:56:08,320 waiting time. 816 00:56:08,320 --> 00:56:10,162 Go ahead. 817 00:56:10,162 --> 00:56:12,632 AUDIENCE: Why is the transfer worth more than 10? 818 00:56:16,315 --> 00:56:18,670 NEEMA NASSIR: That's 10 minutes considered 819 00:56:18,670 --> 00:56:22,020 for one required transfer. 820 00:56:22,020 --> 00:56:24,830 So for example, that's the case where 821 00:56:24,830 --> 00:56:29,090 you have direct access between origin and destination 822 00:56:29,090 --> 00:56:32,450 versus where you need to make at least one transfer to get 823 00:56:32,450 --> 00:56:35,460 to destination. 824 00:56:35,460 --> 00:56:41,970 So the way these coefficients are quantified or understood 825 00:56:41,970 --> 00:56:45,300 is that if you look at the choices 826 00:56:45,300 --> 00:56:49,080 people make in terms of paths that they choose, 827 00:56:49,080 --> 00:56:52,510 given their available path alternatives, 828 00:56:52,510 --> 00:57:00,150 and you can basically calibrate these coefficients based 829 00:57:00,150 --> 00:57:02,160 on the assumption that passengers 830 00:57:02,160 --> 00:57:05,730 try to maximize their utility or minimize 831 00:57:05,730 --> 00:57:08,355 their generalized cost. 832 00:57:13,546 --> 00:57:21,360 There is techniques involving the theory of utility 833 00:57:21,360 --> 00:57:25,720 maximization that is usually used 834 00:57:25,720 --> 00:57:31,530 in practice to look at the actual observed path 835 00:57:31,530 --> 00:57:33,480 choices the passengers are making 836 00:57:33,480 --> 00:57:39,810 and quantify these coefficients. 837 00:57:39,810 --> 00:57:42,860 AUDIENCE: So what if you maximize-- 838 00:57:42,860 --> 00:57:45,680 I mean, for that to result in benefit for customers overall, 839 00:57:45,680 --> 00:57:47,620 you would have to increase, for example, 840 00:57:47,620 --> 00:57:52,660 like headways or frequencies from part of the trip, 841 00:57:52,660 --> 00:57:55,650 but then go back to previous frequency levels? 842 00:57:58,610 --> 00:58:00,920 NEEMA NASSIR: You mean, in the merging? 843 00:58:00,920 --> 00:58:04,940 AUDIENCE: Here, as part of the transfers. 844 00:58:04,940 --> 00:58:09,950 It's only way to maximize the benefit for customers. 845 00:58:09,950 --> 00:58:11,370 NEEMA NASSIR: Right. 846 00:58:11,370 --> 00:58:15,930 So the scenario that I tried to explain here 847 00:58:15,930 --> 00:58:21,900 was removing this route and adding 848 00:58:21,900 --> 00:58:24,990 this layout for the route and replacing it 849 00:58:24,990 --> 00:58:28,300 with this alternative. 850 00:58:28,300 --> 00:58:37,040 So you can keep the same frequency before and after 851 00:58:37,040 --> 00:58:40,110 and the waiting time could be kind of the same. 852 00:58:40,110 --> 00:58:44,370 Or you can alternatively add additional routes 853 00:58:44,370 --> 00:58:49,740 and improve the frequency for other parts of the system 854 00:58:49,740 --> 00:58:55,260 but make sure that you're serving the demand 855 00:58:55,260 --> 00:58:57,210 on the distant suburb. 856 00:58:57,210 --> 00:59:00,120 AUDIENCE: How do you compensate the path right here? 857 00:59:00,120 --> 00:59:01,940 NEEMA NASSIR: This one, right? 858 00:59:01,940 --> 00:59:02,475 This part? 859 00:59:02,475 --> 00:59:04,850 AUDIENCE: Yeah, because they're traveling a longer route. 860 00:59:08,209 --> 00:59:09,000 NEEMA NASSIR: Yeah. 861 00:59:09,000 --> 00:59:11,820 They basically-- the in-vehicle time for these people 862 00:59:11,820 --> 00:59:13,660 will be increased. 863 00:59:13,660 --> 00:59:15,930 So this is the measure like the-- 864 00:59:19,080 --> 00:59:23,850 after you implement this change, if you want to and analyze 865 00:59:23,850 --> 00:59:30,270 the scenario, the in-vehicle time for traveling from-- 866 00:59:30,270 --> 00:59:41,820 it's usually IVT for S1 to D for after will be greater 867 00:59:41,820 --> 00:59:47,810 than IVT for S1 to D, right? 868 00:59:47,810 --> 00:59:52,190 So in-vehicle time will be increased. 869 00:59:52,190 --> 00:59:55,850 However, what's going to change for passengers 870 00:59:55,850 --> 01:00:01,520 who are traveling from S2 to D, the walking distance 871 01:00:01,520 --> 01:00:08,700 is going to decrease, which has a coefficient of two. 872 01:00:08,700 --> 01:00:12,320 The in-vehicle time may increase somehow 873 01:00:12,320 --> 01:00:15,830 because they are boarding now at this point, 874 01:00:15,830 --> 01:00:18,830 but depending on what frequencies you're designing, 875 01:00:18,830 --> 01:00:23,900 you can compute the waiting time and put that into consideration 876 01:00:23,900 --> 01:00:25,130 and into computations. 877 01:00:30,140 --> 01:00:32,090 So how are we doing on time? 878 01:00:34,978 --> 01:00:35,478 OK. 879 01:00:38,870 --> 01:00:46,560 So Toronto financial standards and comparisons-- 880 01:00:46,560 --> 01:00:48,630 basically, service change proposals 881 01:00:48,630 --> 01:00:54,160 are evaluated according to customers gained or lost 882 01:00:54,160 --> 01:00:57,180 per dollar spent or saved. 883 01:00:57,180 --> 01:01:02,130 Again, the benefit that is imagined for transit 884 01:01:02,130 --> 01:01:04,470 is in terms of the number of customers, 885 01:01:04,470 --> 01:01:07,440 the number of passengers that are added and are 886 01:01:07,440 --> 01:01:09,480 using and benefiting from the service. 887 01:01:09,480 --> 01:01:13,500 And the cost side is basically the net cost, 888 01:01:13,500 --> 01:01:17,870 like the operating cost minus the revenue. 889 01:01:17,870 --> 01:01:22,440 And Toronto applies this for new service proposals 890 01:01:22,440 --> 01:01:26,685 and possible reduction to the service and to fare changes. 891 01:01:31,190 --> 01:01:33,540 Again, the financial unit actually 892 01:01:33,540 --> 01:01:36,650 that goes into these computations 893 01:01:36,650 --> 01:01:41,750 is the net cost, operating cost minus the revenue. 894 01:01:41,750 --> 01:01:44,840 And the threshold that is used by Toronto 895 01:01:44,840 --> 01:01:51,530 is 0.23 as the standard threshold, 896 01:01:51,530 --> 01:01:55,700 that's 0.23 passengers per dollar spent. 897 01:01:55,700 --> 01:01:58,220 So for each dollar that you're spending 898 01:01:58,220 --> 01:02:05,490 on the operation minus the revenue, 899 01:02:05,490 --> 01:02:10,700 you need to guarantee at least point 0.23 passengers are 900 01:02:10,700 --> 01:02:12,135 basically getting served. 901 01:02:12,135 --> 01:02:14,422 AUDIENCE: That's operating dollars and passengers. 902 01:02:17,954 --> 01:02:19,862 It's not a capital expense. 903 01:02:19,862 --> 01:02:20,932 It's operating. 904 01:02:20,932 --> 01:02:22,140 NEEMA NASSIR: It's operating. 905 01:02:22,140 --> 01:02:26,760 Yeah, I think it's operating cost minus the revenue, 906 01:02:26,760 --> 01:02:27,820 probably, right. 907 01:02:27,820 --> 01:02:28,800 AUDIENCE: Yeah, OK. 908 01:02:32,230 --> 01:02:34,200 NEEMA NASSIR: Right. 909 01:02:34,200 --> 01:02:38,670 And then services that are operating below this level, 910 01:02:38,670 --> 01:02:43,761 we actually get kind of reviewed in the annual yearly review 911 01:02:43,761 --> 01:02:44,260 process. 912 01:02:48,750 --> 01:02:52,910 Again, in terms of service change process, 913 01:02:52,910 --> 01:02:57,220 major service changes are evaluated twice per year. 914 01:02:57,220 --> 01:03:04,410 They're basically ranked in the comparative analysis 915 01:03:04,410 --> 01:03:09,925 and based on their performance and based on productivity 916 01:03:09,925 --> 01:03:11,280 of the existing service. 917 01:03:13,930 --> 01:03:17,710 There is a board in the Transit Agency 918 01:03:17,710 --> 01:03:19,900 that makes these decisions. 919 01:03:19,900 --> 01:03:29,020 Usually, they invite people from metropolitan transportation 920 01:03:29,020 --> 01:03:34,930 planning agencies too or higher level governmental agencies. 921 01:03:34,930 --> 01:03:37,360 So these boards basically review the recommendations 922 01:03:37,360 --> 01:03:41,020 from the staff and make decisions 923 01:03:41,020 --> 01:03:44,620 based on their merits. 924 01:03:44,620 --> 01:03:48,220 The guiding principles for TransLink 925 01:03:48,220 --> 01:03:51,070 for Vancouver in service optimization program, 926 01:03:51,070 --> 01:03:56,410 is to maintain basic service for transit-dependent customers, 927 01:03:56,410 --> 01:04:01,990 like what we discussed about the existing service 928 01:04:01,990 --> 01:04:04,870 and how passengers may actually be dependent. 929 01:04:04,870 --> 01:04:06,250 And we need to be careful when we 930 01:04:06,250 --> 01:04:08,650 want to remove some of the existing service 931 01:04:08,650 --> 01:04:12,310 and maintain service that are strategic for network 932 01:04:12,310 --> 01:04:14,290 connectivity. 933 01:04:14,290 --> 01:04:16,150 Sometimes it becomes a little tricky 934 01:04:16,150 --> 01:04:21,460 when we look at the routes individually and do not 935 01:04:21,460 --> 01:04:24,970 consider their role and their functionality 936 01:04:24,970 --> 01:04:28,420 in the performance at the network level. 937 01:04:28,420 --> 01:04:32,020 So the guiding principles of TransLink 938 01:04:32,020 --> 01:04:36,250 kind of enforce this concentration. 939 01:04:36,250 --> 01:04:39,880 And then minimize service reductions in area 940 01:04:39,880 --> 01:04:45,820 with no transit alternatives, supporting growing market 941 01:04:45,820 --> 01:04:50,020 where productivity is improving, reinvest in service 942 01:04:50,020 --> 01:04:52,270 to generate higher revenue ridership, 943 01:04:52,270 --> 01:04:57,520 and reinvest in service that supports long-term goals 944 01:04:57,520 --> 01:05:00,820 and objectives of TransLink. 945 01:05:00,820 --> 01:05:06,400 So let's open a discussion now for a few minutes in terms 946 01:05:06,400 --> 01:05:10,930 of possible critiques and possible caveats 947 01:05:10,930 --> 01:05:18,460 of the existing practice for short-range planning. 948 01:05:18,460 --> 01:05:24,490 Let's start-- does anyone have any comments or any thoughts 949 01:05:24,490 --> 01:05:32,490 about the existing caveats or critiques yet? 950 01:05:32,490 --> 01:05:33,540 AUDIENCE: Slide 36. 951 01:05:33,540 --> 01:05:35,390 NEEMA NASSIR: Yeah, it is actually. 952 01:05:35,390 --> 01:05:40,040 AUDIENCE: Where it says the TTC value is, the financial impact 953 01:05:40,040 --> 01:05:41,125 of service changes. 954 01:05:41,125 --> 01:05:43,750 Any service change proposal is evaluated 955 01:05:43,750 --> 01:05:46,940 according to customers gained or lost per dollar spent or saved. 956 01:05:46,940 --> 01:05:49,670 Does not mean that they could eliminate 957 01:05:49,670 --> 01:05:53,210 any stops, for example, and not lose any customers? 958 01:05:53,210 --> 01:05:55,740 They'd just make those customers walk further 959 01:05:55,740 --> 01:05:57,735 and make service less convenient for them, 960 01:05:57,735 --> 01:06:00,730 that that would be a sound decision according 961 01:06:00,730 --> 01:06:03,800 to these criteria? 962 01:06:03,800 --> 01:06:06,260 NEEMA NASSIR: So yeah. 963 01:06:06,260 --> 01:06:12,080 So do you mean, if you're making some changes that reduces 964 01:06:12,080 --> 01:06:14,510 the actual number of customers-- 965 01:06:14,510 --> 01:06:17,600 AUDIENCE: Say they make a change and they don't lose customers 966 01:06:17,600 --> 01:06:21,156 but service is worse for customers. 967 01:06:21,156 --> 01:06:22,910 The customers have no other options, 968 01:06:22,910 --> 01:06:24,858 so they continue using the service, even 969 01:06:24,858 --> 01:06:26,560 thought it's slightly worse. 970 01:06:26,560 --> 01:06:28,617 NEEMA NASSIR: So yeah, that's a good point. 971 01:06:28,617 --> 01:06:29,450 That's a good point. 972 01:06:29,450 --> 01:06:33,320 So probably the important caveats 973 01:06:33,320 --> 01:06:35,960 is that many of these guidelines and standards 974 01:06:35,960 --> 01:06:38,540 are not from the passenger's perspective. 975 01:06:38,540 --> 01:06:42,680 So passenger's perspective performance measures 976 01:06:42,680 --> 01:06:47,080 and passenger's perspective perceptions 977 01:06:47,080 --> 01:06:49,190 of the quality of service is something 978 01:06:49,190 --> 01:06:51,377 that probably should be incorporated 979 01:06:51,377 --> 01:06:52,460 in some of these measures. 980 01:06:56,580 --> 01:06:59,380 AUDIENCE: Isn't it some of the checks and balances in place 981 01:06:59,380 --> 01:07:02,280 because it's relatively easy in services 982 01:07:02,280 --> 01:07:12,056 neutral users of these services lean on it 983 01:07:12,056 --> 01:07:14,170 and some people have to walk further. 984 01:07:14,170 --> 01:07:18,140 So that in itself would lead to a reduction in the number 985 01:07:18,140 --> 01:07:20,630 of people who take the route. 986 01:07:20,630 --> 01:07:23,540 So for example, from suburb S1, the number of people 987 01:07:23,540 --> 01:07:29,920 who [INAUDIBLE] So analysis at route level, because then 988 01:07:29,920 --> 01:07:32,220 it might actually go in favor of the agency. 989 01:07:35,069 --> 01:07:36,110 NEEMA NASSIR: So exactly. 990 01:07:36,110 --> 01:07:39,200 So analysis at the network level should 991 01:07:39,200 --> 01:07:44,150 be something that probably better captured 992 01:07:44,150 --> 01:07:45,630 by these guidelines. 993 01:07:45,630 --> 01:07:51,830 And one of the other things that is not actually included here 994 01:07:51,830 --> 01:07:54,550 is the actual objective of transit agency, which 995 01:07:54,550 --> 01:07:58,520 is improvement and increase of the ridership, which 996 01:07:58,520 --> 01:08:04,190 is not explicitly tied to any of the existing guidelines. 997 01:08:04,190 --> 01:08:08,180 That's something that could be identified thinking of it 998 01:08:08,180 --> 01:08:14,120 as an optimization problem as an optimal or objective function 999 01:08:14,120 --> 01:08:17,649 to be optimized over the solution of space, which 1000 01:08:17,649 --> 01:08:19,460 is the guidelines, existing guidelines 1001 01:08:19,460 --> 01:08:22,460 and standards in the service. 1002 01:08:22,460 --> 01:08:26,840 So what this guideline that we discussed 1003 01:08:26,840 --> 01:08:30,890 here is just the bound on what is acceptable 1004 01:08:30,890 --> 01:08:32,700 and what's not acceptable. 1005 01:08:32,700 --> 01:08:36,649 But the fact that is this going to be beneficial to the system 1006 01:08:36,649 --> 01:08:37,640 or not? 1007 01:08:37,640 --> 01:08:41,540 That's something that requires better analysis of the plan 1008 01:08:41,540 --> 01:08:43,680 basically. 1009 01:08:43,680 --> 01:08:45,960 So any others? 1010 01:08:50,430 --> 01:08:51,154 OK. 1011 01:08:51,154 --> 01:08:53,409 AUDIENCE: Comment on just one thing. 1012 01:08:53,409 --> 01:08:56,810 Is this less for the short term service planning 1013 01:08:56,810 --> 01:08:59,050 and the problem with short-term planning 1014 01:08:59,050 --> 01:09:02,212 often a lot of the demand and [INAUDIBLE] 1015 01:09:02,212 --> 01:09:04,080 will be contemplate. 1016 01:09:04,080 --> 01:09:08,399 Whenever you do major additions, extensions, 1017 01:09:08,399 --> 01:09:10,130 or you have a major budget crisis 1018 01:09:10,130 --> 01:09:12,720 and it's a cut off service, that would not 1019 01:09:12,720 --> 01:09:15,850 be considered entirely a short-term thing 1020 01:09:15,850 --> 01:09:19,622 but more of a bigger impact. 1021 01:09:19,622 --> 01:09:23,735 So the study would be required with demand more. 1022 01:09:23,735 --> 01:09:24,651 NEEMA NASSIR: Correct. 1023 01:09:24,651 --> 01:09:27,020 Exactly. 1024 01:09:27,020 --> 01:09:35,029 OK so in the last slide there is a list of possible caveats 1025 01:09:35,029 --> 01:09:36,170 that are mentioned here. 1026 01:09:36,170 --> 01:09:38,750 Couple of those we already discussed. 1027 01:09:38,750 --> 01:09:43,130 Focus of the existing practice is on poorly-performing routes, 1028 01:09:43,130 --> 01:09:45,710 whereas it does not necessarily have 1029 01:09:45,710 --> 01:09:49,430 to be a poorly-performing route that has the potential 1030 01:09:49,430 --> 01:09:54,380 to be improved to improve the performance of the system. 1031 01:09:54,380 --> 01:09:59,375 There is data limitation both on type and quality. 1032 01:09:59,375 --> 01:10:02,990 We observed that in many of the measurements 1033 01:10:02,990 --> 01:10:05,340 and in many of the decision making, 1034 01:10:05,340 --> 01:10:06,890 there is lots of uncertainty that 1035 01:10:06,890 --> 01:10:11,610 is related to quality of data and limitations of data. 1036 01:10:11,610 --> 01:10:14,180 So hopefully, with the new generations 1037 01:10:14,180 --> 01:10:19,250 of passively-collected data that are nowadays introduced, 1038 01:10:19,250 --> 01:10:25,460 there will be improvement in the existing practice 1039 01:10:25,460 --> 01:10:28,520 to take advantage of that data and improve the planning 1040 01:10:28,520 --> 01:10:30,080 process. 1041 01:10:30,080 --> 01:10:33,000 Again, measures not always closely tied to objectives. 1042 01:10:33,000 --> 01:10:39,700 That's something that is relevant to what 1043 01:10:39,700 --> 01:10:44,570 Adishak mentioned in terms of some of these measures 1044 01:10:44,570 --> 01:10:48,470 are basically just constraints and lower bound 1045 01:10:48,470 --> 01:10:51,920 and they're not necessarily tied to the objective, which 1046 01:10:51,920 --> 01:10:56,780 is improvement of mobility and improvement, increase 1047 01:10:56,780 --> 01:10:58,910 of ridership. 1048 01:10:58,910 --> 01:11:02,090 And then focus on individual routes 1049 01:11:02,090 --> 01:11:06,725 may actually not be ideal because kind of it 1050 01:11:06,725 --> 01:11:10,010 ignores the network level functionality of routes. 1051 01:11:12,660 --> 01:11:14,360 So any questions about this? 1052 01:11:17,490 --> 01:11:19,350 Any additional comments in terms of-- 1053 01:11:22,400 --> 01:11:23,390 right. 1054 01:11:23,390 --> 01:11:24,260 OK. 1055 01:11:24,260 --> 01:11:25,320 Yeah, please. 1056 01:11:25,320 --> 01:11:29,470 AUDIENCE: So the idea of utility is not always something 1057 01:11:29,470 --> 01:11:33,796 that someone actually calculates for when 1058 01:11:33,796 --> 01:11:38,510 they're considering these simulations 1059 01:11:38,510 --> 01:11:42,348 or calculations or changes. 1060 01:11:42,348 --> 01:11:46,300 It's kind of theoretical. 1061 01:11:46,300 --> 01:11:49,060 If we do this then it will affect us 1062 01:11:49,060 --> 01:11:53,927 but there isn't necessarily a quantitative way. 1063 01:11:53,927 --> 01:11:54,760 NEEMA NASSIR: Right. 1064 01:11:54,760 --> 01:12:00,700 The utility itself is a random and unknown quantity, right? 1065 01:12:00,700 --> 01:12:02,530 There's a random component to it. 1066 01:12:02,530 --> 01:12:05,470 But the marginal rate of substitutions 1067 01:12:05,470 --> 01:12:10,560 of these coefficients related to different components, 1068 01:12:10,560 --> 01:12:12,390 that's something that can be quantified. 1069 01:12:12,390 --> 01:12:15,720 For example, what is the disutility 1070 01:12:15,720 --> 01:12:19,740 of waiting time as compared to the disutility 1071 01:12:19,740 --> 01:12:20,970 of in-vehicle time. 1072 01:12:20,970 --> 01:12:27,228 So that's a way kind of have a sense of the generalized cost. 1073 01:12:27,228 --> 01:12:30,216 AUDIENCE: Couldn't you also use a dollar proxy 1074 01:12:30,216 --> 01:12:35,590 to say that the economy is stimulated in x, y, and z? 1075 01:12:35,590 --> 01:12:38,724 Is that another way of looking at it? 1076 01:12:43,030 --> 01:12:48,480 NEEMA NASSIR: I think the-- 1077 01:12:48,480 --> 01:12:50,680 yeah, I'm not sure about-- 1078 01:12:50,680 --> 01:12:53,896 AUDIENCE: It's typically used for capital investments. 1079 01:12:53,896 --> 01:12:55,270 All these people that are looking 1080 01:12:55,270 --> 01:13:00,010 for short-term funds, things like environmental impacts 1081 01:13:00,010 --> 01:13:05,580 and economic welfare at a large scale are equally the methods. 1082 01:13:05,580 --> 01:13:09,690 You are considering a minor extension or increase 1083 01:13:09,690 --> 01:13:13,440 in service with the same alignment. 1084 01:13:13,440 --> 01:13:20,452 So you're probably not going to do a fully [INAUDIBLE] 1085 01:13:20,452 --> 01:13:22,380 gradual marginal change. 1086 01:13:22,380 --> 01:13:25,280 But in a bigger context then, yes, you would have, 1087 01:13:25,280 --> 01:13:27,990 a four-step model and you would look at CO2 1088 01:13:27,990 --> 01:13:31,550 and look at many other things. 1089 01:13:31,550 --> 01:13:34,730 And one example of that would be the recent example 1090 01:13:34,730 --> 01:13:42,270 of London Crossrail, where a lot of the business 1091 01:13:42,270 --> 01:13:50,750 case for Crossrail had to do with deliberation and things 1092 01:13:50,750 --> 01:13:55,360 like this, not necessarily the time that people were saving. 1093 01:14:00,750 --> 01:14:03,185 AUDIENCE: When we say poorly-performing routes, 1094 01:14:03,185 --> 01:14:09,490 are we talking only in terms of timing, for example, 1095 01:14:09,490 --> 01:14:12,640 or are we also talking about poorly-performing 1096 01:14:12,640 --> 01:14:16,887 in terms of financial return of the route? 1097 01:14:16,887 --> 01:14:17,720 NEEMA NASSIR: Right. 1098 01:14:20,360 --> 01:14:21,270 Right. 1099 01:14:21,270 --> 01:14:23,270 AUDIENCE: I'm afraid that for example, if people 1100 01:14:23,270 --> 01:14:30,926 talk about financially, certain routes might get closed 1101 01:14:30,926 --> 01:14:35,347 for low demand, and then you have the equity problem 1102 01:14:35,347 --> 01:14:36,680 that we talked about over there. 1103 01:14:36,680 --> 01:14:37,250 NEEMA NASSIR: Right. 1104 01:14:37,250 --> 01:14:37,770 Yeah. 1105 01:14:37,770 --> 01:14:43,340 So I believe when you're evaluating 1106 01:14:43,340 --> 01:14:46,280 the quality of these routes, we can 1107 01:14:46,280 --> 01:14:49,340 use all the existing criteria that we mentioned. 1108 01:14:49,340 --> 01:14:53,090 For example, the criteria that is related to access, 1109 01:14:53,090 --> 01:14:56,870 the criteria that is related to span of the service, access 1110 01:14:56,870 --> 01:15:00,470 over time and space, and then measures 1111 01:15:00,470 --> 01:15:04,100 that are dealing with the productivity of the route 1112 01:15:04,100 --> 01:15:07,760 and the financial aspects. 1113 01:15:07,760 --> 01:15:14,300 So my guess is that this is probably 1114 01:15:14,300 --> 01:15:16,940 kind of when we're evaluating the routes, 1115 01:15:16,940 --> 01:15:20,060 we are taking into account the equity issue 1116 01:15:20,060 --> 01:15:24,800 and the access issue and all those sorts of guidelines 1117 01:15:24,800 --> 01:15:29,450 that we discussed last week into consideration.