1 00:00:01,040 --> 00:00:03,380 The following content is provided under a Creative 2 00:00:03,380 --> 00:00:04,770 Commons license. 3 00:00:04,770 --> 00:00:06,980 Your support will help MIT OpenCourseWare 4 00:00:06,980 --> 00:00:11,070 continue to offer high-quality educational resources for free. 5 00:00:11,070 --> 00:00:13,640 To make a donation or to view additional materials 6 00:00:13,640 --> 00:00:17,600 from hundreds of MIT courses, visit MIT OpenCourseWare 7 00:00:17,600 --> 00:00:18,486 at ocw.mit.edu. 8 00:00:22,154 --> 00:00:24,710 NEEMA NASSIR: Let's go ahead and start. 9 00:00:24,710 --> 00:00:26,730 Good afternoon, everyone. 10 00:00:26,730 --> 00:00:28,070 My name is Neema Nassir. 11 00:00:28,070 --> 00:00:31,520 I'm a post-doctoral associate at Transit Lab, MIT. 12 00:00:31,520 --> 00:00:36,110 I've been invited to give a few lectures in this course 13 00:00:36,110 --> 00:00:41,180 and basically participate in delivering the course 14 00:00:41,180 --> 00:00:42,630 material for this course. 15 00:00:42,630 --> 00:00:47,970 I will be talking in this session 16 00:00:47,970 --> 00:00:52,130 and in the next session on Short-range Planning Practice. 17 00:00:52,130 --> 00:00:57,950 And then I'll be back again in April and in four sessions 18 00:00:57,950 --> 00:01:01,670 in April, lecture number 13, 14, 15, and 16 19 00:01:01,670 --> 00:01:04,410 dealing with vehicle scheduling, crew scheduling, network 20 00:01:04,410 --> 00:01:11,840 and root structure and high ridership corridors. 21 00:01:11,840 --> 00:01:15,090 So let's go ahead and start. 22 00:01:15,090 --> 00:01:20,070 In this session, we are going to talk about short-range planning 23 00:01:20,070 --> 00:01:21,420 practice. 24 00:01:21,420 --> 00:01:25,890 And basically, I will give a brief set of definition 25 00:01:25,890 --> 00:01:29,400 and concepts and introduction to the planning 26 00:01:29,400 --> 00:01:32,310 activities that are usually made at transit authorities 27 00:01:32,310 --> 00:01:37,150 in the US and also worldwide. 28 00:01:37,150 --> 00:01:40,200 And then we will briefly talk about the measures 29 00:01:40,200 --> 00:01:44,550 and the standards and guidelines that are published and used 30 00:01:44,550 --> 00:01:51,690 by transit agencies to plan their service based on those. 31 00:01:51,690 --> 00:01:56,820 Then we will also talk about current practice and critique 32 00:01:56,820 --> 00:02:00,780 of the existing systems. 33 00:02:00,780 --> 00:02:06,630 Probably we will cover icon 1 and some material in icon 2 34 00:02:06,630 --> 00:02:07,650 this week. 35 00:02:07,650 --> 00:02:13,230 And then the rest will actually be delivered to you 36 00:02:13,230 --> 00:02:14,865 in our session next Tuesday. 37 00:02:20,570 --> 00:02:26,350 So the planning decisions and the planning activities that 38 00:02:26,350 --> 00:02:30,740 are usually done at the transportation planning 39 00:02:30,740 --> 00:02:34,490 agencies and public transport planning agencies usually can 40 00:02:34,490 --> 00:02:38,990 be classified into four different type of decisions 41 00:02:38,990 --> 00:02:42,500 depending on frequency of decision-making and the need 42 00:02:42,500 --> 00:02:47,840 for making decisions and the scope of decisions-- 43 00:02:47,840 --> 00:02:52,580 the scope of changes that are related to the decisions that 44 00:02:52,580 --> 00:02:55,370 are made with each activity. 45 00:02:55,370 --> 00:02:57,530 These planning activities are divided 46 00:02:57,530 --> 00:03:01,130 into four different classes-- 47 00:03:01,130 --> 00:03:05,540 long range planning activities, medium range, short range, 48 00:03:05,540 --> 00:03:07,640 and control. 49 00:03:07,640 --> 00:03:11,480 Long range planning activities basically include 50 00:03:11,480 --> 00:03:13,490 major capital investment-- 51 00:03:13,490 --> 00:03:16,370 major improvements to the infrastructure-- 52 00:03:16,370 --> 00:03:21,230 which would include construction of a new station, a new rail 53 00:03:21,230 --> 00:03:24,050 line, a new bus way to the system. 54 00:03:24,050 --> 00:03:28,820 Or it could also be related to and as a result 55 00:03:28,820 --> 00:03:36,660 of major institutional changes and changes to the, basically, 56 00:03:36,660 --> 00:03:41,180 financial resources of the organization. 57 00:03:41,180 --> 00:03:44,420 The decisions that [INAUDIBLE] involve long range planning 58 00:03:44,420 --> 00:03:49,730 are usually driven by political and economic considerations. 59 00:03:49,730 --> 00:03:53,510 And basically, the type of analysis 60 00:03:53,510 --> 00:03:57,470 and the type of planning that is done in long range planning 61 00:03:57,470 --> 00:04:03,740 activities involve projection into the future and prediction 62 00:04:03,740 --> 00:04:07,860 of demand and prediction of situation in the future. 63 00:04:07,860 --> 00:04:10,760 Therefore, it requires a comprehensive analysis 64 00:04:10,760 --> 00:04:16,940 usually for any strategic planning of interactions 65 00:04:16,940 --> 00:04:20,120 between changes in land use and interaction 66 00:04:20,120 --> 00:04:22,460 between land use and demand in the future. 67 00:04:22,460 --> 00:04:27,290 And it basically needs to capture or have, like, 68 00:04:27,290 --> 00:04:34,470 an insight into the future of the network and of the region. 69 00:04:34,470 --> 00:04:38,700 The next range is basically or the next class 70 00:04:38,700 --> 00:04:42,750 of planning activities is basically 71 00:04:42,750 --> 00:04:45,000 the medium range activities. 72 00:04:45,000 --> 00:04:49,080 Medium range activities include bus network structure design, 73 00:04:49,080 --> 00:04:53,190 network size design and improvement, fleet size, 74 00:04:53,190 --> 00:04:55,020 fare policy and technology. 75 00:04:55,020 --> 00:05:01,590 It basically involves incremental or medium 76 00:05:01,590 --> 00:05:08,865 improvement to the system and to the transit service. 77 00:05:11,510 --> 00:05:14,750 One point that I need to mention about the long range and medium 78 00:05:14,750 --> 00:05:17,900 range is that these two, although 79 00:05:17,900 --> 00:05:20,900 are really important in the planning 80 00:05:20,900 --> 00:05:23,330 practice for transportation systems, 81 00:05:23,330 --> 00:05:28,400 but these decisions are usually done at the upper level-- 82 00:05:28,400 --> 00:05:32,090 at the higher lever than the transit authority. 83 00:05:32,090 --> 00:05:34,190 For example, long range decisions 84 00:05:34,190 --> 00:05:39,890 are usually done in collaboration 85 00:05:39,890 --> 00:05:43,700 between NPO's metropolitan planning organizations, 86 00:05:43,700 --> 00:05:45,530 state departments of transportation, 87 00:05:45,530 --> 00:05:46,910 and transit authorities. 88 00:05:46,910 --> 00:05:49,490 So the transit authority is not the final decision maker 89 00:05:49,490 --> 00:05:50,630 in that case. 90 00:05:50,630 --> 00:05:55,940 And it needs close interaction and close collaboration 91 00:05:55,940 --> 00:05:58,190 with the higher level organizations 92 00:05:58,190 --> 00:06:01,309 and governmental agencies to make the final decisions. 93 00:06:01,309 --> 00:06:02,600 So that's about the long range. 94 00:06:02,600 --> 00:06:04,760 Medium range is also somehow similar. 95 00:06:04,760 --> 00:06:06,500 It usually, for example, the decisions 96 00:06:06,500 --> 00:06:09,740 that are related to the network structure, although there 97 00:06:09,740 --> 00:06:15,050 is a lot of interesting research going on in universities that 98 00:06:15,050 --> 00:06:18,680 deals with optimization of network design and optimization 99 00:06:18,680 --> 00:06:22,760 of the structure of the network, because what happens 100 00:06:22,760 --> 00:06:26,750 in practice is that usually in the networks that we live 101 00:06:26,750 --> 00:06:30,830 and in current cities, the actual layout and structure 102 00:06:30,830 --> 00:06:32,760 of the network is already given. 103 00:06:32,760 --> 00:06:37,300 And it's usually rare or impossible that we 104 00:06:37,300 --> 00:06:41,630 start from scratch and start to design a network from scratch 105 00:06:41,630 --> 00:06:44,490 to begin with. 106 00:06:44,490 --> 00:06:47,660 So therefore, what we usually observe 107 00:06:47,660 --> 00:06:51,620 is incremental changes to the network and addition 108 00:06:51,620 --> 00:06:56,210 or elimination of some routes from the service and basically, 109 00:06:56,210 --> 00:06:58,184 decisions that are related to that. 110 00:06:58,184 --> 00:06:59,492 AUDIENCE: I have a question. 111 00:06:59,492 --> 00:07:00,575 NEEMA NASSIR: Yes, please. 112 00:07:00,575 --> 00:07:02,700 AUDIENCE: In terms of fleet size, 113 00:07:02,700 --> 00:07:04,710 could that come as a request from the agencies? 114 00:07:04,710 --> 00:07:07,320 Like, an agency says, we have a bus line. 115 00:07:07,320 --> 00:07:09,200 It's at capacity or over capacity. 116 00:07:09,200 --> 00:07:10,790 We need more buses. 117 00:07:10,790 --> 00:07:13,670 Could it come, like, from a request to the higher level? 118 00:07:13,670 --> 00:07:15,020 NEEMA NASSIR: Yes, definitely. 119 00:07:15,020 --> 00:07:23,060 Yeah, but it eventually requires the interaction on those, 120 00:07:23,060 --> 00:07:27,380 so securing the funding from the upper level 121 00:07:27,380 --> 00:07:30,010 governmental agencies-- 122 00:07:30,010 --> 00:07:31,820 so yeah. 123 00:07:31,820 --> 00:07:36,440 So the reason that I'm explaining this 124 00:07:36,440 --> 00:07:41,000 is that I'd like for you to understand why we are focusing 125 00:07:41,000 --> 00:07:44,030 on short range decision-making and short range planning 126 00:07:44,030 --> 00:07:46,130 practice in this course. 127 00:07:46,130 --> 00:07:50,480 That's mainly because the bulk part of activities at transit 128 00:07:50,480 --> 00:07:55,490 agencies is related to short range planning and short range 129 00:07:55,490 --> 00:07:57,430 decision-making. 130 00:07:57,430 --> 00:08:00,990 So what we mean by short range decision-making 131 00:08:00,990 --> 00:08:04,890 are decisions that are usually made more frequently 132 00:08:04,890 --> 00:08:07,560 than a yearly basis. 133 00:08:07,560 --> 00:08:10,830 And they basically entail decisions 134 00:08:10,830 --> 00:08:13,740 about the structures of the route-- 135 00:08:13,740 --> 00:08:15,990 layout of the route-- layout of the stops-- 136 00:08:15,990 --> 00:08:22,380 and basically decisions about the frequency of service-- 137 00:08:22,380 --> 00:08:26,520 timetable-- and vehicle and crew scheduling. 138 00:08:26,520 --> 00:08:33,210 Despite the top two classes of decision-making 139 00:08:33,210 --> 00:08:39,659 that are usually driven by political and economic 140 00:08:39,659 --> 00:08:42,210 considerations, short range decision making 141 00:08:42,210 --> 00:08:49,890 and control activities are usually basically driven 142 00:08:49,890 --> 00:08:52,110 by cost optimization. 143 00:08:52,110 --> 00:08:55,590 And these are basically tactical decisions as opposed 144 00:08:55,590 --> 00:08:58,620 to strategic decisions that are usually in the long range 145 00:08:58,620 --> 00:09:00,170 and medium range. 146 00:09:00,170 --> 00:09:07,960 So then there's real time and control decision-making, 147 00:09:07,960 --> 00:09:11,100 which is basically usually done currently 148 00:09:11,100 --> 00:09:16,890 in using expert decision and expert activities. 149 00:09:20,940 --> 00:09:26,790 There is an ongoing effort in the technology 150 00:09:26,790 --> 00:09:29,580 to improve the quality of control-- 151 00:09:29,580 --> 00:09:32,850 ongoing research to use real time data 152 00:09:32,850 --> 00:09:36,060 and improve the real time performance 153 00:09:36,060 --> 00:09:38,070 by using real time control. 154 00:09:38,070 --> 00:09:43,380 This includes revisions of the routes. 155 00:09:43,380 --> 00:09:48,300 That could be the case in reaction 156 00:09:48,300 --> 00:09:51,390 to incidents and disruption situation. 157 00:09:51,390 --> 00:09:55,960 It could also relate to holding and dispatching decisions-- 158 00:09:55,960 --> 00:09:56,460 thank you-- 159 00:09:59,250 --> 00:10:02,700 that could actually be used to improve 160 00:10:02,700 --> 00:10:11,550 the effective capacity of the system by trying to regulate-- 161 00:10:11,550 --> 00:10:18,730 have a better frequency, basically, in the service. 162 00:10:18,730 --> 00:10:26,470 So similar to the typical planning problem, 163 00:10:26,470 --> 00:10:28,780 there are major elements in planning 164 00:10:28,780 --> 00:10:31,750 for public transportation and particularly 165 00:10:31,750 --> 00:10:37,930 in short term and short range decision-making. 166 00:10:37,930 --> 00:10:41,380 It includes data collection similar to what 167 00:10:41,380 --> 00:10:46,750 you guys have done in your homework assignment number 2. 168 00:10:46,750 --> 00:10:49,660 Collection of data on demand, collection of data on supply, 169 00:10:49,660 --> 00:10:51,790 and interaction of supply on demand 170 00:10:51,790 --> 00:10:55,660 is an essential part to be able to make decisions and make them 171 00:10:55,660 --> 00:11:00,040 basically reliable and actionable intelligence 172 00:11:00,040 --> 00:11:02,740 for the decision-making process. 173 00:11:02,740 --> 00:11:07,630 Then there should be analysis on identification 174 00:11:07,630 --> 00:11:10,960 of potential problems and identification of opportunities 175 00:11:10,960 --> 00:11:13,210 to improve the situation. 176 00:11:13,210 --> 00:11:16,900 And then there is the creative part 177 00:11:16,900 --> 00:11:20,050 of design of options, solution, and strategies 178 00:11:20,050 --> 00:11:22,000 to address the situation. 179 00:11:22,000 --> 00:11:28,180 And then typical to any economic analysis, 180 00:11:28,180 --> 00:11:30,880 there would be an element of cost 181 00:11:30,880 --> 00:11:37,450 and benefit estimation to be able to justify the project. 182 00:11:40,180 --> 00:11:43,810 So in terms of operational planning process 183 00:11:43,810 --> 00:11:49,030 or in another term short range planning process, 184 00:11:49,030 --> 00:11:52,450 there are multiple stages involved in the work. 185 00:11:52,450 --> 00:11:58,450 And usually, what happens is that in practice, 186 00:11:58,450 --> 00:12:02,050 it is proven that the planning procedure 187 00:12:02,050 --> 00:12:06,580 can be decomposed into sequential decision-making 188 00:12:06,580 --> 00:12:08,080 processes. 189 00:12:08,080 --> 00:12:14,860 And although it is possible to formulate 190 00:12:14,860 --> 00:12:18,850 all of these decisions into one gigantic optimization problem 191 00:12:18,850 --> 00:12:22,210 and solve it for all the decision variables 192 00:12:22,210 --> 00:12:25,910 that you need to basically address, however, 193 00:12:25,910 --> 00:12:28,180 the practice has shown that there's 194 00:12:28,180 --> 00:12:31,300 a nice way of decomposing the process 195 00:12:31,300 --> 00:12:35,170 and to solve problems and solve them sequentially in order 196 00:12:35,170 --> 00:12:40,390 to make sure that you're satisfying 197 00:12:40,390 --> 00:12:44,660 all the constraints to the accurate level 198 00:12:44,660 --> 00:12:47,290 and you're gaining the-- 199 00:12:47,290 --> 00:12:50,050 you finally get a realistic optimal solution 200 00:12:50,050 --> 00:12:54,700 for your planning practice. 201 00:12:54,700 --> 00:13:02,850 Considerations regarding policy constraints and some 202 00:13:02,850 --> 00:13:05,500 of the economic concerns are usually 203 00:13:05,500 --> 00:13:10,630 difficult to formulate into optimization problems. 204 00:13:10,630 --> 00:13:13,090 But there are published guidelines 205 00:13:13,090 --> 00:13:17,590 by governmental agencies that you can basically 206 00:13:17,590 --> 00:13:20,740 follow and make sure that at least the top two 207 00:13:20,740 --> 00:13:25,390 level of the decision-making processes and tasks 208 00:13:25,390 --> 00:13:30,710 here are basically done to the satisfactory level. 209 00:13:30,710 --> 00:13:33,760 So let's go ahead and start and review these tasks. 210 00:13:33,760 --> 00:13:35,380 There are lectures in this course that 211 00:13:35,380 --> 00:13:39,730 are dealing with each one of these tasks 212 00:13:39,730 --> 00:13:42,610 and components in planning. 213 00:13:42,610 --> 00:13:46,060 But in this lecture, I'll just briefly 214 00:13:46,060 --> 00:13:48,850 go over these different components 215 00:13:48,850 --> 00:13:56,000 as an introductory part of the course. 216 00:13:56,000 --> 00:14:02,460 So for every optimization problem 217 00:14:02,460 --> 00:14:06,435 or in this case planning process, 218 00:14:06,435 --> 00:14:08,040 we have a set of inputs-- 219 00:14:08,040 --> 00:14:12,390 constraints-- basically, that you need to satisfy. 220 00:14:12,390 --> 00:14:14,790 And then there's an objective function 221 00:14:14,790 --> 00:14:17,880 that you need to optimize-- so inputs. 222 00:14:17,880 --> 00:14:24,840 So the first component of the planning and practice 223 00:14:24,840 --> 00:14:28,060 here is bus route design. 224 00:14:28,060 --> 00:14:31,860 This includes design of layout of the routes 225 00:14:31,860 --> 00:14:35,700 and layout of the stops. 226 00:14:35,700 --> 00:14:39,030 And basically, the input into this task 227 00:14:39,030 --> 00:14:44,040 includes constraints that are related to right-of-way roadway 228 00:14:44,040 --> 00:14:47,640 constraints, like if there is a river in your network, 229 00:14:47,640 --> 00:14:48,900 there's this connectivity. 230 00:14:48,900 --> 00:14:50,580 There are bridges-- tunnels. 231 00:14:50,580 --> 00:14:54,120 So these are the kind of constraints at network level 232 00:14:54,120 --> 00:14:56,310 that you would like to satisfy. 233 00:14:56,310 --> 00:14:58,980 And there's also demand constraints in the network 234 00:14:58,980 --> 00:15:02,380 that you need to consider for this task, 235 00:15:02,380 --> 00:15:07,140 like, you have a sense of what's the demand from certain suburb 236 00:15:07,140 --> 00:15:09,500 to the downtown area. 237 00:15:09,500 --> 00:15:12,450 And you basically would like to have a route 238 00:15:12,450 --> 00:15:16,470 system that kind of can address that demand 239 00:15:16,470 --> 00:15:18,090 and satisfy that demand. 240 00:15:18,090 --> 00:15:23,521 Then the solution to the design component and that task 241 00:15:23,521 --> 00:15:25,020 [INAUDIBLE] out of routes and layout 242 00:15:25,020 --> 00:15:28,650 of the stops that are basically resulting 243 00:15:28,650 --> 00:15:32,440 from the first component, and then 244 00:15:32,440 --> 00:15:34,980 the set of routes and stops will be 245 00:15:34,980 --> 00:15:40,560 input into the second task, which is setting timetables. 246 00:15:40,560 --> 00:15:45,150 This task includes finding frequencies for the routes 247 00:15:45,150 --> 00:15:50,040 and from that identifying and defining-- 248 00:15:50,040 --> 00:15:52,260 basically, generating the timetables 249 00:15:52,260 --> 00:15:54,410 based on frequencies. 250 00:15:54,410 --> 00:15:59,250 There are usually considerations related to policy headways, 251 00:15:59,250 --> 00:16:02,850 policy frequencies, and capacity frequencies 252 00:16:02,850 --> 00:16:07,240 that need to be addressed in this task. 253 00:16:07,240 --> 00:16:12,660 But again, this task is also somehow dealt 254 00:16:12,660 --> 00:16:15,450 with using the existing guidelines 255 00:16:15,450 --> 00:16:18,380 and just satisfying these guidelines. 256 00:16:18,380 --> 00:16:20,400 The input and other constraints that 257 00:16:20,400 --> 00:16:22,920 are considered for this task would 258 00:16:22,920 --> 00:16:25,680 be level of service guidelines, and again, 259 00:16:25,680 --> 00:16:30,360 the demand constraints in the network. 260 00:16:30,360 --> 00:16:35,940 Then from the set of timetables, what will be resulted here 261 00:16:35,940 --> 00:16:40,530 is the departure times that was scheduled for the service. 262 00:16:40,530 --> 00:16:45,900 So the first two components of row 1 and row 2, 263 00:16:45,900 --> 00:16:49,580 basically, are dealing with satisfying the demand 264 00:16:49,580 --> 00:16:51,280 and satisfying the guidelines. 265 00:16:51,280 --> 00:16:55,600 But once you go beyond the first two components, 266 00:16:55,600 --> 00:16:57,914 it becomes to the level that you know 267 00:16:57,914 --> 00:17:01,050 kind of what's the service you want to deliver. 268 00:17:01,050 --> 00:17:03,150 And then you're dealing with tactical decision-- 269 00:17:03,150 --> 00:17:08,579 how you can optimize the cost of delivering this service 270 00:17:08,579 --> 00:17:10,180 and improve efficiency. 271 00:17:10,180 --> 00:17:13,589 So did the third row here basically deals 272 00:17:13,589 --> 00:17:17,369 with the departure times that are 273 00:17:17,369 --> 00:17:19,470 given for all the routes in the network 274 00:17:19,470 --> 00:17:22,500 and for all times of day. 275 00:17:22,500 --> 00:17:24,180 And then what you need to do is how 276 00:17:24,180 --> 00:17:28,079 you want to allocate your resources in terms 277 00:17:28,079 --> 00:17:33,210 of allocate the vehicles to the system 278 00:17:33,210 --> 00:17:35,940 to make sure that that service is 279 00:17:35,940 --> 00:17:41,190 going to be delivered on time. 280 00:17:41,190 --> 00:17:45,680 So that would be the task of vehicle scheduling input 281 00:17:45,680 --> 00:17:47,830 into that task or constraints that are involved 282 00:17:47,830 --> 00:17:54,080 or running times on different segments of the route 283 00:17:54,080 --> 00:17:57,930 and also the departure times that are given from upper level 284 00:17:57,930 --> 00:17:59,580 decision-making process. 285 00:17:59,580 --> 00:18:02,730 And it will eventually result into a schedule 286 00:18:02,730 --> 00:18:07,020 of given buses in terms of what are specific trips they 287 00:18:07,020 --> 00:18:11,020 need to serve and what's the sequence of these trips. 288 00:18:11,020 --> 00:18:16,660 Then the output from that decision with the vehicle 289 00:18:16,660 --> 00:18:22,450 schedules into the task of scheduling for drivers, 290 00:18:22,450 --> 00:18:26,890 this task also include input from operator 291 00:18:26,890 --> 00:18:31,950 and basically work union constraints, 292 00:18:31,950 --> 00:18:37,540 like maximum shift hours and, for example, like, 293 00:18:37,540 --> 00:18:40,870 break times and lunchtimes and maximum separation 294 00:18:40,870 --> 00:18:42,130 between two shifts. 295 00:18:42,130 --> 00:18:44,050 And there are rules and regulations 296 00:18:44,050 --> 00:18:45,430 that are related to that. 297 00:18:45,430 --> 00:18:50,770 And this task, again, is a task of allocating resources 298 00:18:50,770 --> 00:18:55,240 to minimize the cost of scheduling drivers 299 00:18:55,240 --> 00:18:58,000 to deliver the service. 300 00:18:58,000 --> 00:19:00,520 And the result would be the schedules for the crew. 301 00:19:03,720 --> 00:19:09,450 So then these decisions can be looked at in another way, 302 00:19:09,450 --> 00:19:13,170 too, like, we can further classify these decisions 303 00:19:13,170 --> 00:19:16,470 into service planning decisions and operational planning 304 00:19:16,470 --> 00:19:18,030 decision. 305 00:19:18,030 --> 00:19:20,880 Service planning decisions would be the decisions 306 00:19:20,880 --> 00:19:23,790 that define service as it is perceived and understood 307 00:19:23,790 --> 00:19:26,700 by public. 308 00:19:26,700 --> 00:19:29,160 This is basically the part of service 309 00:19:29,160 --> 00:19:32,435 that is visible to the public. 310 00:19:32,435 --> 00:19:33,810 And then there are decisions that 311 00:19:33,810 --> 00:19:36,660 are related to operations, which basically 312 00:19:36,660 --> 00:19:40,290 defines how operations occur to produce the service. 313 00:19:40,290 --> 00:19:46,770 This is somehow a black box to the passengers. 314 00:19:46,770 --> 00:19:51,390 And yeah, basically, the passengers 315 00:19:51,390 --> 00:19:55,140 should not really care what's the vehicle's schedule 316 00:19:55,140 --> 00:19:58,950 and what's the crew's schedule that is assigned to the trip, 317 00:19:58,950 --> 00:20:02,572 as long as trip is being delivered as it is scheduled. 318 00:20:05,410 --> 00:20:09,100 So in terms of other characteristics 319 00:20:09,100 --> 00:20:12,670 of these decisions, we, basically, 320 00:20:12,670 --> 00:20:18,790 for network design frequency setting time table development, 321 00:20:18,790 --> 00:20:22,320 vehicle scheduling, and crew scheduling, 322 00:20:22,320 --> 00:20:28,570 the decisions that are involved in planning, basically, 323 00:20:28,570 --> 00:20:31,840 starting from top to bottom frequency of decisions, 324 00:20:31,840 --> 00:20:35,140 basically are increased. 325 00:20:35,140 --> 00:20:40,570 And the scope of decision-making into the future 326 00:20:40,570 --> 00:20:44,545 is basically become limited and smaller. 327 00:20:47,530 --> 00:20:55,360 And also, the decision-making in the network design frequency 328 00:20:55,360 --> 00:20:59,110 setting and timetabling is basically 329 00:20:59,110 --> 00:21:02,650 made based on guidelines and standards. 330 00:21:02,650 --> 00:21:06,700 And decision-making on the lower level-- vehicle scheduling 331 00:21:06,700 --> 00:21:09,220 and crew scheduling-- is made usually 332 00:21:09,220 --> 00:21:11,650 based on optimization of cost. 333 00:21:11,650 --> 00:21:18,230 And then given the nature of the decision-making, the tools that 334 00:21:18,230 --> 00:21:20,480 are used and the techniques that are 335 00:21:20,480 --> 00:21:25,150 used for solving these planning decision-making processes 336 00:21:25,150 --> 00:21:33,160 are basically ranging from judgments and manual analysis 337 00:21:33,160 --> 00:21:36,250 that is more dominated into computer tools that 338 00:21:36,250 --> 00:21:40,420 are basically designed to solve optimization problems 339 00:21:40,420 --> 00:21:43,000 for minimizing the cost. 340 00:21:45,750 --> 00:21:52,080 So similar to the classical evaluation structure, 341 00:21:52,080 --> 00:21:56,910 when you're dealing with planning questions and planning 342 00:21:56,910 --> 00:22:01,400 projects or planning alternatives, 343 00:22:01,400 --> 00:22:04,950 there's this structure that starts with goals, objectives, 344 00:22:04,950 --> 00:22:07,740 measures, and standards or guidelines. 345 00:22:07,740 --> 00:22:15,300 So let's just make a quick practice and define a goal 346 00:22:15,300 --> 00:22:17,520 and then define objectives for that goal, 347 00:22:17,520 --> 00:22:20,010 and define measures based on that, 348 00:22:20,010 --> 00:22:32,170 and see how a standard can be defined to make that planning 349 00:22:32,170 --> 00:22:37,130 decision acceptable. 350 00:22:37,130 --> 00:22:40,920 So let's define a goal. 351 00:22:49,110 --> 00:22:51,180 Does anyone have any suggestion? 352 00:22:53,790 --> 00:22:58,648 What is an important goal for public transportation system? 353 00:22:58,648 --> 00:23:00,256 AUDIENCE: Maximize ridership. 354 00:23:00,256 --> 00:23:02,870 NEEMA NASSIR: Maximize ridership, yes. 355 00:23:02,870 --> 00:23:10,020 So let's say you want to maximize ridership. 356 00:23:10,020 --> 00:23:11,520 That's your research problem, right? 357 00:23:11,520 --> 00:23:12,000 AUDIENCE: [CHUCKLING] 358 00:23:12,000 --> 00:23:12,480 NEEMA NASSIR: [CHUCKLING] 359 00:23:12,480 --> 00:23:13,440 AUDIENCE: [INAUDIBLE] 360 00:23:13,440 --> 00:23:14,330 NEEMA NASSIR: Right. 361 00:23:14,330 --> 00:23:19,810 [LAUGHTER] So then in terms of objectives, 362 00:23:19,810 --> 00:23:22,330 what would be an objective that we can define for this? 363 00:23:25,860 --> 00:23:31,830 I would say probably the peak hour demand 364 00:23:31,830 --> 00:23:34,380 could be one objective or peak hour 365 00:23:34,380 --> 00:23:36,660 ridership in certain corridor. 366 00:23:36,660 --> 00:23:47,600 Let's say peak hour ridership per weekday per month. 367 00:23:56,625 --> 00:24:00,100 Well, peak hour ridership would be the objective here. 368 00:24:00,100 --> 00:24:01,850 And then the measure that you would define 369 00:24:01,850 --> 00:24:06,170 would be peak hour ridership per weekday per month. 370 00:24:06,170 --> 00:24:16,310 And then the standard that you want to define 371 00:24:16,310 --> 00:24:22,670 would be something that not only captures 372 00:24:22,670 --> 00:24:26,090 the demand side, but also the supply of the system, too. 373 00:24:26,090 --> 00:24:30,445 So you would come up with probably, like-- 374 00:24:30,445 --> 00:24:31,070 I would, like-- 375 00:24:31,070 --> 00:24:32,320 AUDIENCE: Riders per vehicle-- 376 00:24:32,320 --> 00:24:35,570 riders per mile-- per hour. 377 00:24:35,570 --> 00:24:36,840 NEEMA NASSIR: Yeah, exactly. 378 00:24:36,840 --> 00:24:41,210 Or let's say, like, 60% of capacity. 379 00:24:41,210 --> 00:24:47,160 Like, your standard in this case would 380 00:24:47,160 --> 00:24:54,010 be your peak hour ridership should be greater than 60% 381 00:24:54,010 --> 00:24:57,020 of capacity. 382 00:24:57,020 --> 00:24:58,590 So this was just an example that we 383 00:24:58,590 --> 00:25:03,150 tried to work out one practice for evaluation structure. 384 00:25:03,150 --> 00:25:05,430 But let me give you another example that is actually 385 00:25:05,430 --> 00:25:10,030 used in the practice. 386 00:25:12,980 --> 00:25:14,900 So assume that we are considering 387 00:25:14,900 --> 00:25:20,250 the goal of reliability in the transit system. 388 00:25:20,250 --> 00:25:25,130 The goal of reliability would be considered 389 00:25:25,130 --> 00:25:32,570 one of the main goals because if we 390 00:25:32,570 --> 00:25:37,850 have a reliable public transport system, you can secure 391 00:25:37,850 --> 00:25:42,140 and you can guarantee a body of demand 392 00:25:42,140 --> 00:25:46,970 in your network that can rely on your transit system 393 00:25:46,970 --> 00:25:56,740 and basically adapt to the lifestyle of using 394 00:25:56,740 --> 00:25:57,830 public transport. 395 00:25:57,830 --> 00:26:03,160 And that's beneficial to the system 396 00:26:03,160 --> 00:26:07,160 from different perspectives. 397 00:26:07,160 --> 00:26:10,450 So the objective that we can define 398 00:26:10,450 --> 00:26:15,712 for reliability in this case would be possibly-- 399 00:26:15,712 --> 00:26:17,170 can you come up with some objective 400 00:26:17,170 --> 00:26:22,210 that we can define with the goal of improving reliability? 401 00:26:22,210 --> 00:26:23,650 AUDIENCE: On-time performance. 402 00:26:23,650 --> 00:26:25,275 NEEMA NASSIR: On-time performance, yes. 403 00:26:30,340 --> 00:26:34,940 So on-time performance would be the objective. 404 00:26:34,940 --> 00:26:38,850 And then we need to define a measure for this 405 00:26:38,850 --> 00:26:42,690 to be able to measure that objective 406 00:26:42,690 --> 00:26:46,680 for each one of the scenarios and for each one 407 00:26:46,680 --> 00:26:50,440 of the alternative plans. 408 00:26:50,440 --> 00:26:55,112 So what would be a measure for on-time performance? 409 00:26:55,112 --> 00:27:01,407 AUDIENCE: Characterizing with our car driving [INAUDIBLE] 410 00:27:01,407 --> 00:27:02,740 within a certain amount of time. 411 00:27:02,740 --> 00:27:03,823 NEEMA NASSIR: Right, yeah. 412 00:27:03,823 --> 00:27:10,680 So let's say percentage of on-time arrivals. 413 00:27:13,660 --> 00:27:17,130 So we would say percentage of on-time arrivals, 414 00:27:17,130 --> 00:27:19,470 and we would define on-time arrivals 415 00:27:19,470 --> 00:27:24,470 to be between 0 and 5 minutes of the schedule. 416 00:27:27,770 --> 00:27:31,900 And then the standard that many of the agencies 417 00:27:31,900 --> 00:27:39,400 use to measure this would be 90% or 95%, basically. 418 00:27:46,010 --> 00:27:52,020 So this way, the transit agency can 419 00:27:52,020 --> 00:27:57,330 go ahead and evaluate the performance of the operators 420 00:27:57,330 --> 00:28:00,810 of the system who basically are running 421 00:28:00,810 --> 00:28:04,260 some part of the service and come back to them 422 00:28:04,260 --> 00:28:08,550 and say, OK, your performance in terms of reliability 423 00:28:08,550 --> 00:28:10,220 has been acceptable or not. 424 00:28:10,220 --> 00:28:10,720 Please. 425 00:28:10,720 --> 00:28:15,753 AUDIENCE: [INAUDIBLE] something [INAUDIBLE].. 426 00:28:15,753 --> 00:28:22,891 Or would it be just standard operating procdedure 427 00:28:22,891 --> 00:28:24,655 they were planning as well. 428 00:28:24,655 --> 00:28:27,450 And their reliability was somewhere around 80%, 429 00:28:27,450 --> 00:28:30,127 and now the goal is to achieve a 10% gain. 430 00:28:30,127 --> 00:28:30,960 NEEMA NASSIR: Mm-hm. 431 00:28:30,960 --> 00:28:33,460 AUDIENCE: And comparing that with the previous one would 432 00:28:33,460 --> 00:28:34,774 be a benchmark again. 433 00:28:34,774 --> 00:28:35,940 NEEMA NASSIR: Yeah, exactly. 434 00:28:35,940 --> 00:28:41,610 So another measure would be percentage improvements as 435 00:28:41,610 --> 00:28:43,320 compared to baseline. 436 00:28:43,320 --> 00:28:46,790 And then the transit agency can define a standard 437 00:28:46,790 --> 00:28:50,160 of at least 10% improvement to the baseline 438 00:28:50,160 --> 00:28:53,130 should be required for reliability. 439 00:28:53,130 --> 00:28:54,840 Yep, thank you. 440 00:28:54,840 --> 00:29:04,290 So then one other interesting point 441 00:29:04,290 --> 00:29:07,620 that actually comes to mind looking at the existing 442 00:29:07,620 --> 00:29:10,860 guidelines is that on-time arrivals are usually 443 00:29:10,860 --> 00:29:13,980 defined for 0 to 5. 444 00:29:13,980 --> 00:29:17,220 That basically means that you consider 445 00:29:17,220 --> 00:29:21,290 a bus that is late by two minutes to be on time. 446 00:29:21,290 --> 00:29:23,520 But a bus that is early by one minute 447 00:29:23,520 --> 00:29:25,360 is not considered to be on time. 448 00:29:25,360 --> 00:29:30,990 Does anyone have any comments about this-- 449 00:29:30,990 --> 00:29:34,550 why we do not consider early arrivals on time? 450 00:29:34,550 --> 00:29:37,612 AUDIENCE: I think misconnections for people 451 00:29:37,612 --> 00:29:39,877 who expected to be there at a certain time, but-- 452 00:29:39,877 --> 00:29:40,960 NEEMA NASSIR: Right, yeah. 453 00:29:40,960 --> 00:29:41,820 AUDIENCE: --they left. 454 00:29:41,820 --> 00:29:42,330 NEEMA NASSIR: Yeah, yeah. 455 00:29:42,330 --> 00:29:43,663 AUDIENCE: [INAUDIBLE] connected. 456 00:29:43,663 --> 00:29:45,040 NEEMA NASSIR: So it can be proved 457 00:29:45,040 --> 00:29:51,330 that based on typical assumptions about passenger 458 00:29:51,330 --> 00:29:57,120 arrivals, early arrivals of buses and service 459 00:29:57,120 --> 00:30:02,000 is usually more negative as later arrivals. 460 00:30:02,000 --> 00:30:03,540 Let's make an example here. 461 00:30:03,540 --> 00:30:09,960 For example, let's assume under uniform arrivals 462 00:30:09,960 --> 00:30:11,490 of passengers-- 463 00:30:11,490 --> 00:30:14,070 let's assume if this is my scheduled 464 00:30:14,070 --> 00:30:17,790 arrivals over the axis of time and this 465 00:30:17,790 --> 00:30:24,480 is my number of passengers waiting in the bus stop 466 00:30:24,480 --> 00:30:30,900 under uniform arrivals, [INAUDIBLE] 467 00:30:30,900 --> 00:30:35,520 you have some accumulation of passengers 468 00:30:35,520 --> 00:30:36,810 that are waiting at the stop. 469 00:30:42,880 --> 00:30:48,370 And then what happens is that if one of these buses are early-- 470 00:30:48,370 --> 00:30:53,120 let's say it basically arrives at this time-- 471 00:30:53,120 --> 00:30:55,350 what happens is that the amount of time 472 00:30:55,350 --> 00:30:59,500 that we are saving due to this arrival 473 00:30:59,500 --> 00:31:03,670 is basically less than the amount of time that 474 00:31:03,670 --> 00:31:09,700 will be added to our waiting time for the next trip. 475 00:31:09,700 --> 00:31:12,200 And that's because the number of people who are experiencing 476 00:31:12,200 --> 00:31:15,140 this is a little higher-- 477 00:31:15,140 --> 00:31:17,080 number of people that experience it, 478 00:31:17,080 --> 00:31:21,110 saying, the additional waiting time 479 00:31:21,110 --> 00:31:23,300 is higher than number of people who are 480 00:31:23,300 --> 00:31:27,410 experiencing the early arrival. 481 00:31:27,410 --> 00:31:29,960 And it cuts to their waiting time. 482 00:31:29,960 --> 00:31:39,130 So if you compare these two areas, that basically tells you 483 00:31:39,130 --> 00:31:42,240 you're basically having a loss or having 484 00:31:42,240 --> 00:31:46,360 an increase in waiting time if your service is early. 485 00:31:46,360 --> 00:31:49,210 And then the situation becomes worse-- 486 00:31:49,210 --> 00:31:52,920 indicates that you suggested-- when people are coordinating 487 00:31:52,920 --> 00:31:54,550 with the system-- 488 00:31:54,550 --> 00:31:57,240 for example, it is assumed that-- 489 00:31:57,240 --> 00:32:02,340 and it can actually be proven by looking at the data, too-- 490 00:32:02,340 --> 00:32:06,420 people make the transition to coordinated behavior 491 00:32:06,420 --> 00:32:10,740 and coordination with the schedule as the frequency 492 00:32:10,740 --> 00:32:13,110 of service increases. 493 00:32:13,110 --> 00:32:17,640 For example, for a service that is scheduled 494 00:32:17,640 --> 00:32:22,380 to be delivered every half hour or every one hour, 495 00:32:22,380 --> 00:32:24,810 it is expected that passengers are 496 00:32:24,810 --> 00:32:28,995 coordinating with the system and coordinating 497 00:32:28,995 --> 00:32:34,500 their arrivals with the departure of the buses. 498 00:32:34,500 --> 00:32:39,450 But when it comes to more frequent service, 499 00:32:39,450 --> 00:32:45,537 then we will have more uniform arrivals. 500 00:32:45,537 --> 00:32:47,120 So in that case, what we would observe 501 00:32:47,120 --> 00:32:49,490 is that passengers arrivals would resemble something 502 00:32:49,490 --> 00:32:50,175 like this. 503 00:32:55,400 --> 00:33:01,390 Well, it has to go up to here, the original case. 504 00:33:01,390 --> 00:33:07,690 And then in that case if your service early by five 505 00:33:07,690 --> 00:33:13,150 or 10 minutes, you will have the main bulk of your passengers 506 00:33:13,150 --> 00:33:17,500 who need to wait another headway, basically. 507 00:33:17,500 --> 00:33:21,820 So this should be actually added to your computation, too. 508 00:33:24,114 --> 00:33:25,530 Is there any questions about this? 509 00:33:34,660 --> 00:33:43,150 So in terms of the advantage of guidelines in transit planning, 510 00:33:43,150 --> 00:33:46,230 one of the main purposes of these guidelines 511 00:33:46,230 --> 00:33:48,930 is to communicate to the public the rationale 512 00:33:48,930 --> 00:33:51,510 and the logic behind your decision-making 513 00:33:51,510 --> 00:33:54,990 and behind your allocation of resources to the system. 514 00:33:54,990 --> 00:33:58,470 It also gives you a consistent and fair basis 515 00:33:58,470 --> 00:34:00,360 for decision-making. 516 00:34:00,360 --> 00:34:07,260 And it is somehow gives you a tool to justify your decisions 517 00:34:07,260 --> 00:34:10,320 and also a tool to make sure that your decisions are 518 00:34:10,320 --> 00:34:13,830 fair and justified in nature. 519 00:34:13,830 --> 00:34:16,949 It also gives you the opportunity 520 00:34:16,949 --> 00:34:22,080 to balance the improvement and the investments to the network 521 00:34:22,080 --> 00:34:26,040 and gain a uniform level of service improvement 522 00:34:26,040 --> 00:34:29,620 for efficient service. 523 00:34:29,620 --> 00:34:34,030 So the guidelines that are existing in the literature 524 00:34:34,030 --> 00:34:37,570 and in the practice, basically-- in the state of practice-- 525 00:34:37,570 --> 00:34:42,940 are dealing with the design of service 526 00:34:42,940 --> 00:34:44,860 and with the delivery of service. 527 00:34:44,860 --> 00:34:49,360 So the design of service basically dealing 528 00:34:49,360 --> 00:34:54,489 with variables that are related to spacing 529 00:34:54,489 --> 00:34:58,360 and density of stops, spacing and density of trips, 530 00:34:58,360 --> 00:35:02,410 and frequency of the service as it is designed and as it 531 00:35:02,410 --> 00:35:03,830 is in the schedule. 532 00:35:03,830 --> 00:35:05,500 So there are guidelines related to that. 533 00:35:05,500 --> 00:35:09,160 And there is also guidelines related to how you deliver 534 00:35:09,160 --> 00:35:12,820 or how the transit agency delivers the design service 535 00:35:12,820 --> 00:35:18,340 and how the performance actually took place. 536 00:35:18,340 --> 00:35:24,570 So the main factors of service design and service quality 537 00:35:24,570 --> 00:35:28,590 from the passenger's perspective are basically 538 00:35:28,590 --> 00:35:31,620 reported by transit capacity and quality 539 00:35:31,620 --> 00:35:38,340 of service manual of the Transportation Research Board. 540 00:35:38,340 --> 00:35:45,360 And it basically includes frequency of the service as one 541 00:35:45,360 --> 00:35:51,090 of the main aspects or factors that are defining 542 00:35:51,090 --> 00:35:52,410 the quality of service-- 543 00:35:52,410 --> 00:35:55,050 waiting time, reliability, and access. 544 00:35:55,050 --> 00:35:59,130 This is access over time and also access over space. 545 00:35:59,130 --> 00:36:02,970 So why do we again have both frequency and waiting 546 00:36:02,970 --> 00:36:04,990 time in here? 547 00:36:04,990 --> 00:36:11,340 So if frequency and waiting time are kind of interrelated, 548 00:36:11,340 --> 00:36:17,460 how come there are two separate items here on the list here? 549 00:36:17,460 --> 00:36:20,070 Does anyone have any suggestions about this? 550 00:36:31,860 --> 00:36:37,990 So let me go ahead and explain how I understand this. 551 00:36:37,990 --> 00:36:43,600 Frequency of the service is important not only 552 00:36:43,600 --> 00:36:45,550 in terms of the waiting time-- 553 00:36:45,550 --> 00:36:47,500 that could be associated with that-- 554 00:36:47,500 --> 00:36:52,041 but also when it comes to lower frequency system and longer 555 00:36:52,041 --> 00:36:52,540 headways. 556 00:36:55,310 --> 00:36:58,360 The waiting time is not a function of frequency anymore. 557 00:36:58,360 --> 00:37:04,010 It basically, as we discussed, the waiting time 558 00:37:04,010 --> 00:37:08,450 could actually be independent of frequency. 559 00:37:08,450 --> 00:37:12,280 However, in those cases, there is a burden of coordination 560 00:37:12,280 --> 00:37:14,900 that is imposed to the passenger. 561 00:37:14,900 --> 00:37:19,060 So if you have to coordinate your arrival, for example, 562 00:37:19,060 --> 00:37:21,280 in the morning from home to work, 563 00:37:21,280 --> 00:37:24,190 then there is some sort of inconvenience 564 00:37:24,190 --> 00:37:29,740 that is imposed to you to basically make it 565 00:37:29,740 --> 00:37:32,980 to your coordinated schedule. 566 00:37:32,980 --> 00:37:34,930 And then waiting time is also important 567 00:37:34,930 --> 00:37:41,100 because if the service is frequent 568 00:37:41,100 --> 00:37:45,640 but at different ranges of frequency, 569 00:37:45,640 --> 00:37:47,800 you may experience different waiting time 570 00:37:47,800 --> 00:37:51,830 if the process of arrivals of passenger 571 00:37:51,830 --> 00:37:55,060 is assumed to be a random process. 572 00:37:55,060 --> 00:37:57,340 Then reliability as we discussed is 573 00:37:57,340 --> 00:38:00,940 one of the important features and one 574 00:38:00,940 --> 00:38:04,350 of the important aspects of the service. 575 00:38:04,350 --> 00:38:08,230 It usually is a measure based on on-time arrivals 576 00:38:08,230 --> 00:38:09,430 of the service. 577 00:38:09,430 --> 00:38:16,340 And then the next aspect that is important 578 00:38:16,340 --> 00:38:20,630 is access to origins and destinations. 579 00:38:20,630 --> 00:38:24,560 This access can be interpreted in two ways. 580 00:38:24,560 --> 00:38:27,820 We can look at access in terms of space and in terms of time. 581 00:38:27,820 --> 00:38:32,290 Access in terms of space is closeness to the bus stops 582 00:38:32,290 --> 00:38:36,490 and closeness to the service from origin 583 00:38:36,490 --> 00:38:37,930 and from the destination. 584 00:38:37,930 --> 00:38:40,810 Access in terms of time would be the span of service-- 585 00:38:40,810 --> 00:38:43,000 when you like to travel-- 586 00:38:43,000 --> 00:38:45,125 and basically, what is the service time 587 00:38:45,125 --> 00:38:48,115 and available times of service. 588 00:38:51,260 --> 00:38:53,390 So yeah, as I mentioned, most agencies 589 00:38:53,390 --> 00:38:56,000 have guidelines covering a span of service. 590 00:38:56,000 --> 00:39:01,250 For example, MBTA has this guideline 591 00:39:01,250 --> 00:39:07,850 that for weekdays, you need to have service from all suburbs 592 00:39:07,850 --> 00:39:13,130 into the CBD area with-- 593 00:39:13,130 --> 00:39:16,910 at least, like, at latest, arrives at 7:00 AM. 594 00:39:16,910 --> 00:39:21,620 So you need to basically be able to cover between 7:00 595 00:39:21,620 --> 00:39:28,760 AM and 6:00 PM as your services span in MBTA. 596 00:39:28,760 --> 00:39:33,080 For TransLink, it's also somehow the same. 597 00:39:33,080 --> 00:39:35,750 You need to-- basically, minimum service guidelines 598 00:39:35,750 --> 00:39:38,810 to ensure that 95% of trips listed 599 00:39:38,810 --> 00:39:43,620 can be completed in the time span given below. 600 00:39:43,620 --> 00:39:47,210 Like, from any point to downtown Vancouver, 601 00:39:47,210 --> 00:39:51,980 you need to have service that can actually 602 00:39:51,980 --> 00:39:54,170 arrive in downtown Vancouver by 7:00 603 00:39:54,170 --> 00:39:59,090 AM, for people who should start their work at 7:00 AM. 604 00:39:59,090 --> 00:40:03,890 And then there's also trips in the evening. 605 00:40:03,890 --> 00:40:09,110 And it basically says earliest departure time of mass transit 606 00:40:09,110 --> 00:40:11,840 trip in the evening from downtown Vancouver 607 00:40:11,840 --> 00:40:15,320 to any town center needs to be midnight for people 608 00:40:15,320 --> 00:40:17,540 who basically would like to enjoy 609 00:40:17,540 --> 00:40:21,090 the nightlife in the downtown area 610 00:40:21,090 --> 00:40:23,180 and then use transit to go back home. 611 00:40:26,560 --> 00:40:29,740 So in terms of route design and in terms 612 00:40:29,740 --> 00:40:34,480 of layout of routes and stops, there 613 00:40:34,480 --> 00:40:38,290 are also guidelines that are addressing this decision. 614 00:40:38,290 --> 00:40:41,410 MBTA has a policy objective to provide transit service 615 00:40:41,410 --> 00:40:45,070 within walking distance, which is defined a quarter 616 00:40:45,070 --> 00:40:48,760 mile of all residents that are living in area with population 617 00:40:48,760 --> 00:40:53,230 densities greater than 5,000 people per square mile. 618 00:40:53,230 --> 00:40:58,150 That's basically an interesting guideline. 619 00:40:58,150 --> 00:41:00,860 It considers two different variables. 620 00:41:00,860 --> 00:41:02,530 One is the population density. 621 00:41:02,530 --> 00:41:07,420 And the other one is the density of the stops or vicinity 622 00:41:07,420 --> 00:41:12,850 or walkability of transit in areas 623 00:41:12,850 --> 00:41:17,290 that are populated with densities greater than 5,000 624 00:41:17,290 --> 00:41:20,440 per square mile. 625 00:41:20,440 --> 00:41:25,530 So in terms of comprehensiveness objective, 626 00:41:25,530 --> 00:41:30,030 TransLink basically has very comprehensive guidelines 627 00:41:30,030 --> 00:41:36,900 related to basically a set of different performance measures 628 00:41:36,900 --> 00:41:38,590 and goals. 629 00:41:38,590 --> 00:41:43,260 Let me read from these. 630 00:41:43,260 --> 00:41:46,410 At least 90% of our residents and employees 631 00:41:46,410 --> 00:41:48,120 in urbanized development areas should 632 00:41:48,120 --> 00:41:52,500 have a walk less than 450 meters to a bus stop. 633 00:41:52,500 --> 00:41:57,420 So this is related to access over space. 634 00:41:57,420 --> 00:41:58,275 98%-- 635 00:41:58,275 --> 00:42:01,350 AUDIENCE: When they say 90% of residents and employees, 636 00:42:01,350 --> 00:42:05,010 basically they're looking at that person twice. 637 00:42:05,010 --> 00:42:07,150 The person is on one end, a resident, 638 00:42:07,150 --> 00:42:08,640 and on the other end, an employee. 639 00:42:08,640 --> 00:42:10,200 NEEMA NASSIR: Right, yes. 640 00:42:10,200 --> 00:42:11,235 AUDIENCE: So it's-- OK. 641 00:42:11,235 --> 00:42:15,740 NEEMA NASSIR: Yeah, you want to consider both ends of the trip, 642 00:42:15,740 --> 00:42:16,380 right? 643 00:42:16,380 --> 00:42:19,110 And these percentages are a little tricky. 644 00:42:19,110 --> 00:42:21,540 Sometimes they're counting the number of people-- number 645 00:42:21,540 --> 00:42:22,170 of residents. 646 00:42:22,170 --> 00:42:25,200 Sometimes they're counting OD pairs. 647 00:42:25,200 --> 00:42:27,180 Sometimes they're counting passengers. 648 00:42:27,180 --> 00:42:33,270 So it's a little tricky to interpret these correctly. 649 00:42:33,270 --> 00:42:37,350 So the second one basically says 98% of all peak period transit 650 00:42:37,350 --> 00:42:39,720 trips to and from downtown Vancouver 651 00:42:39,720 --> 00:42:43,560 should require no more than one transfer. 652 00:42:43,560 --> 00:42:48,930 That basically is 98% of people who are traveling from 653 00:42:48,930 --> 00:42:51,430 and to downtown to the suburbs. 654 00:42:51,430 --> 00:42:55,800 And so it may be a little difficult, especially 655 00:42:55,800 --> 00:42:59,700 in the absence of AFC data-- 656 00:42:59,700 --> 00:43:02,610 Automated Fare Collection system data. 657 00:43:02,610 --> 00:43:08,270 It can be a little tricky to measure these objectives. 658 00:43:08,270 --> 00:43:10,890 What's usually done, basically, is 659 00:43:10,890 --> 00:43:16,980 people go out and basically ask from transit riders 660 00:43:16,980 --> 00:43:18,720 about their origins and destinations. 661 00:43:18,720 --> 00:43:21,510 It's called Public Transit Origin Destination Survey 662 00:43:21,510 --> 00:43:23,530 that is usually done on board. 663 00:43:23,530 --> 00:43:24,874 Do you have a question? 664 00:43:24,874 --> 00:43:25,374 OK. 665 00:43:28,330 --> 00:43:31,600 So the second measure here relates 666 00:43:31,600 --> 00:43:34,820 to connectivity in the network. 667 00:43:34,820 --> 00:43:38,290 The third one is 95% of all peak period 668 00:43:38,290 --> 00:43:40,990 transit trips to the nearest town centers 669 00:43:40,990 --> 00:43:42,980 should require no more than one transfer. 670 00:43:42,980 --> 00:43:47,630 Again, this one is also related to connectivity of the network. 671 00:43:47,630 --> 00:43:51,670 And fourth one is all trips between one town center 672 00:43:51,670 --> 00:43:55,270 and adjacent town centers should require no transfers. 673 00:43:55,270 --> 00:43:59,530 Again, this one is a guideline on connectivity. 674 00:43:59,530 --> 00:44:04,390 And then again, the last one is also on connectivity. 675 00:44:04,390 --> 00:44:09,160 It basically requires that 95% of all peak period transit 676 00:44:09,160 --> 00:44:13,300 trips to major regional activity centers and gateways 677 00:44:13,300 --> 00:44:15,820 should require no more than two transfers. 678 00:44:19,090 --> 00:44:20,940 OK, so any questions or comments about this? 679 00:44:20,940 --> 00:44:21,120 AUDIENCE: Yeah. 680 00:44:21,120 --> 00:44:21,911 NEEMA NASSIR: Yeah. 681 00:44:21,911 --> 00:44:24,370 AUDIENCE: In terms of one connectivity between one town 682 00:44:24,370 --> 00:44:28,290 center and adjacent town center, definition of a town 683 00:44:28,290 --> 00:44:29,550 varies from place to place. 684 00:44:29,550 --> 00:44:29,960 NEEMA NASSIR: Mm-hm. 685 00:44:29,960 --> 00:44:31,830 AUDIENCE: I know that specifically Vancouver 686 00:44:31,830 --> 00:44:34,770 has rather large suburbs-- 687 00:44:34,770 --> 00:44:36,240 like, suburbs of 100,000 people-- 688 00:44:36,240 --> 00:44:37,230 200,000 people. 689 00:44:37,230 --> 00:44:41,770 So you actually have a few municipal entities 690 00:44:41,770 --> 00:44:42,479 for a large area. 691 00:44:42,479 --> 00:44:43,311 NEEMA NASSIR: Mm-hm. 692 00:44:43,311 --> 00:44:46,110 AUDIENCE: But you can have areas where that condition would 693 00:44:46,110 --> 00:44:48,900 be a little bit too strict if you have, let's say, 694 00:44:48,900 --> 00:44:51,030 the Boston area, which is generally 695 00:44:51,030 --> 00:44:53,760 slightly smaller suburbs and more suburbs to connect. 696 00:44:53,760 --> 00:44:55,500 NEEMA NASSIR: Right, yeah. 697 00:44:55,500 --> 00:44:58,380 So some of these may actually be a little tricky 698 00:44:58,380 --> 00:45:00,765 to measure and to quantify. 699 00:45:00,765 --> 00:45:01,490 AUDIENCE: Yeah. 700 00:45:01,490 --> 00:45:02,656 And they're able to compare. 701 00:45:02,656 --> 00:45:03,660 NEEMA NASSIR: Right. 702 00:45:03,660 --> 00:45:06,680 AUDIENCE: Yeah, when I see these numbers, then I 703 00:45:06,680 --> 00:45:11,592 try to think about equity, and, like, the 10% that is not 450-- 704 00:45:11,592 --> 00:45:12,425 NEEMA NASSIR: Right. 705 00:45:12,425 --> 00:45:13,360 AUDIENCE: --is it from that or-- 706 00:45:13,360 --> 00:45:13,910 NEEMA NASSIR: Yeah. 707 00:45:13,910 --> 00:45:14,400 AUDIENCE: --a bus station-- 708 00:45:14,400 --> 00:45:14,730 NEEMA NASSIR: Right. 709 00:45:14,730 --> 00:45:15,630 AUDIENCE: --[INAUDIBLE] the 5% . 710 00:45:15,630 --> 00:45:16,464 NEEMA NASSIR: Mm-hm. 711 00:45:16,464 --> 00:45:18,880 AUDIENCE: So how do you navigate these types of trade-offs 712 00:45:18,880 --> 00:45:21,565 in the face of equity. 713 00:45:21,565 --> 00:45:22,690 NEEMA NASSIR: Right. 714 00:45:22,690 --> 00:45:24,910 This is a really good point. 715 00:45:24,910 --> 00:45:27,220 Some of these percentages, I think, 716 00:45:27,220 --> 00:45:31,180 are related to random variables. 717 00:45:31,180 --> 00:45:34,360 And those are somehow acceptable because that's 718 00:45:34,360 --> 00:45:37,570 when you try to define a confidence interval. 719 00:45:37,570 --> 00:45:41,170 But when it comes to cases like this, 720 00:45:41,170 --> 00:45:43,490 it is not a random variable anymore. 721 00:45:43,490 --> 00:45:50,470 It's basically like a special variable that constantly 722 00:45:50,470 --> 00:45:55,270 requires a certain amount of connections to the downtown 723 00:45:55,270 --> 00:45:56,380 area, right? 724 00:45:56,380 --> 00:46:00,760 So that may be one of the loopholes of this guideline. 725 00:46:00,760 --> 00:46:04,490 That's a really good point because one 726 00:46:04,490 --> 00:46:08,440 may argue that if 95% of my OD pairs 727 00:46:08,440 --> 00:46:12,160 are connected with the minimum connectivity requirement, 728 00:46:12,160 --> 00:46:14,230 what about the rest of 5%? 729 00:46:14,230 --> 00:46:18,160 Does it mean that all people who live in that 5% area 730 00:46:18,160 --> 00:46:23,950 will have to go through the unacceptable connectivity 731 00:46:23,950 --> 00:46:25,360 situation? 732 00:46:25,360 --> 00:46:26,530 So that's a valid point. 733 00:46:26,530 --> 00:46:28,450 And thank you for bringing that up. 734 00:46:28,450 --> 00:46:30,910 AUDIENCE: Does Vancouver actually achieve these goals? 735 00:46:30,910 --> 00:46:34,814 It seems kind of implausible-- kind of difficult to-- 736 00:46:34,814 --> 00:46:35,790 NEEMA NASSIR: Right. 737 00:46:35,790 --> 00:46:36,936 AUDIENCE: [INAUDIBLE] 738 00:46:36,936 --> 00:46:41,790 NEEMA NASSIR: Yeah, well, these are basically usually 739 00:46:41,790 --> 00:46:47,160 for projects and plans that are being proposed 740 00:46:47,160 --> 00:46:50,190 for-- or like small start projects 741 00:46:50,190 --> 00:46:53,470 or for improvement plans. 742 00:46:53,470 --> 00:46:57,120 So I think what they try to do is to make sure 743 00:46:57,120 --> 00:47:01,740 that the ones that are being accepted and being implemented 744 00:47:01,740 --> 00:47:03,330 would satisfy this. 745 00:47:03,330 --> 00:47:06,960 But I'm not really sure if the existing network satisfy 746 00:47:06,960 --> 00:47:08,370 all these conditions or not. 747 00:47:08,370 --> 00:47:09,407 AUDIENCE: Yes. 748 00:47:09,407 --> 00:47:11,502 AUDIENCE: Sorry, is TransLink the name 749 00:47:11,502 --> 00:47:12,710 of the agency that runs the-- 750 00:47:12,710 --> 00:47:13,850 NEEMA NASSIR: Yes, right. 751 00:47:13,850 --> 00:47:16,800 Yeah, right. 752 00:47:16,800 --> 00:47:19,680 OK, so any other questions? 753 00:47:19,680 --> 00:47:20,660 All right. 754 00:47:26,550 --> 00:47:31,300 So then TransLink guideline on route design 755 00:47:31,300 --> 00:47:32,520 is very interesting, too. 756 00:47:32,520 --> 00:47:35,880 They have really interesting logic 757 00:47:35,880 --> 00:47:40,770 to validate the design and the improvements that 758 00:47:40,770 --> 00:47:43,290 are proposed for route design. 759 00:47:43,290 --> 00:47:46,710 So what it says is that deviations from the most direct 760 00:47:46,710 --> 00:47:50,760 route must have walking time savings for customers 761 00:47:50,760 --> 00:47:52,560 on the added route section greater 762 00:47:52,560 --> 00:47:55,590 than the increase in total travel time 763 00:47:55,590 --> 00:47:57,120 for through passengers. 764 00:47:57,120 --> 00:48:03,820 So does anyone have any interpretation-- 765 00:48:03,820 --> 00:48:08,455 quick interpretation-- on this guideline? 766 00:48:08,455 --> 00:48:11,800 AUDIENCE: The only reason is [INAUDIBLE] a major destination 767 00:48:11,800 --> 00:48:14,710 where people [INAUDIBLE] or otherwise people 768 00:48:14,710 --> 00:48:20,500 are going straight and what [INAUDIBLE] people on buses. 769 00:48:20,500 --> 00:48:22,940 So the second to the last stop on the route, 770 00:48:22,940 --> 00:48:24,800 may be you make that deviation to have 771 00:48:24,800 --> 00:48:26,890 people on the bus [INAUDIBLE] traffic [INAUDIBLE] 772 00:48:26,890 --> 00:48:27,760 NEEMA NASSIR: Right. 773 00:48:27,760 --> 00:48:28,760 Yeah. 774 00:48:28,760 --> 00:48:29,340 Right. 775 00:48:29,340 --> 00:48:30,339 You have a comment, too? 776 00:48:30,339 --> 00:48:31,876 AUDIENCE: [INAUDIBLE] 777 00:48:31,876 --> 00:48:33,440 NEEMA NASSIR: Right, yeah. 778 00:48:33,440 --> 00:48:39,110 So if you look at the problem or look at the situation 779 00:48:39,110 --> 00:48:41,630 from the perspective of system and from the perspective 780 00:48:41,630 --> 00:48:45,860 of total travel time in the system, what they basically 781 00:48:45,860 --> 00:48:48,590 are requiring is that, for example, 782 00:48:48,590 --> 00:48:51,260 let's assume that I have a suburb here. 783 00:48:51,260 --> 00:48:53,330 And then this is my downtown area. 784 00:48:53,330 --> 00:48:56,810 And there is a route here that has multiple stops. 785 00:48:56,810 --> 00:48:59,600 And then there's another suburb here, 786 00:48:59,600 --> 00:49:03,350 which, in the original plan is being served by this route. 787 00:49:03,350 --> 00:49:08,990 However, it requires a walking link to different stops. 788 00:49:08,990 --> 00:49:14,090 What this guideline basically requires is that if the total 789 00:49:14,090 --> 00:49:18,050 travel time for all passengers for traveling from here to here 790 00:49:18,050 --> 00:49:21,770 and for all that are traveling from here to here is going 791 00:49:21,770 --> 00:49:29,870 to be increased, then you cannot justify an additional detour 792 00:49:29,870 --> 00:49:34,670 here to serve this area, meaning that if the total travel time 793 00:49:34,670 --> 00:49:41,930 saving that is done on the walking time of these people is 794 00:49:41,930 --> 00:49:45,890 not as much as total increase in travel time of people that are 795 00:49:45,890 --> 00:49:49,440 traveling from this suburb to this suburb, 796 00:49:49,440 --> 00:49:53,160 then you cannot justify the design or the re-route. 797 00:50:00,150 --> 00:50:03,810 So in terms of guidelines related 798 00:50:03,810 --> 00:50:07,050 to a schedule and guidelines related to frequency 799 00:50:07,050 --> 00:50:11,070 and timetable, we basically, in practice there 800 00:50:11,070 --> 00:50:14,790 are two main components-- component 801 00:50:14,790 --> 00:50:20,220 that is related to the policy of design and components 802 00:50:20,220 --> 00:50:22,260 that is related to the crowding situation. 803 00:50:22,260 --> 00:50:26,190 Policy of design basically dictate a minimum headway 804 00:50:26,190 --> 00:50:30,300 or basically a minimal frequency or maximum 805 00:50:30,300 --> 00:50:33,510 headway for the service. 806 00:50:33,510 --> 00:50:36,940 And that's basically the case when 807 00:50:36,940 --> 00:50:38,670 there is no crowding situation. 808 00:50:41,590 --> 00:50:44,640 However, in the main corridors and in peak directions, 809 00:50:44,640 --> 00:50:50,400 we usually would go beyond the service of policy headway 810 00:50:50,400 --> 00:50:54,780 because we need to accommodate the demand. 811 00:50:54,780 --> 00:51:02,070 And that's basically how the maximum passenger crowding 812 00:51:02,070 --> 00:51:03,830 guides come into play. 813 00:51:06,510 --> 00:51:13,890 So in terms of the actual guideline on policy of headway 814 00:51:13,890 --> 00:51:16,530 for MBTA, we have maximum headway 815 00:51:16,530 --> 00:51:19,710 on all local bus routes should be 30 minutes in peak 816 00:51:19,710 --> 00:51:21,360 and 60 minutes at other times. 817 00:51:21,360 --> 00:51:23,880 For express routes, there should be at least three trips 818 00:51:23,880 --> 00:51:24,820 in each peak period. 819 00:51:27,880 --> 00:51:31,720 So let's discuss a little bit more about this. 820 00:51:31,720 --> 00:51:35,340 Why do you think that we need to have a policy headway-- 821 00:51:35,340 --> 00:51:37,000 maximum policy headway? 822 00:51:39,800 --> 00:51:43,610 If we are considering the demand and we 823 00:51:43,610 --> 00:51:47,900 are trying to satisfy the demand by number of vehicles 824 00:51:47,900 --> 00:51:52,130 that we are assigning to the route, 825 00:51:52,130 --> 00:51:54,541 then why should we care about a minimum? 826 00:51:54,541 --> 00:51:55,040 Do you-- 827 00:51:55,040 --> 00:51:59,315 AUDIENCE: Because if the demand is really low, 828 00:51:59,315 --> 00:52:01,729 that person-- you can't make him wait forever. 829 00:52:01,729 --> 00:52:02,520 NEEMA NASSIR: Yeah. 830 00:52:02,520 --> 00:52:03,765 AUDIENCE: You'll have to serve him equitably. 831 00:52:03,765 --> 00:52:05,280 NEEMA NASSIR: Exactly, yeah. 832 00:52:05,280 --> 00:52:09,775 Again, the equity issue and the accessibility issue, basically. 833 00:52:09,775 --> 00:52:11,400 AUDIENCE: And there are also, you know, 834 00:52:11,400 --> 00:52:14,520 still elderly people who don't use smartphones 835 00:52:14,520 --> 00:52:16,590 to coordinate their arrival at the bus 836 00:52:16,590 --> 00:52:17,930 stop when the bus is coming. 837 00:52:17,930 --> 00:52:18,540 NEEMA NASSIR: Right. 838 00:52:18,540 --> 00:52:21,040 AUDIENCE: And you don't want to make them wait indefinitely. 839 00:52:21,040 --> 00:52:23,050 NEEMA NASSIR: Right, exactly. 840 00:52:23,050 --> 00:52:25,550 Right. 841 00:52:25,550 --> 00:52:30,460 OK, and then when it comes to considerations related 842 00:52:30,460 --> 00:52:32,680 to the crowding, we have guidelines 843 00:52:32,680 --> 00:52:35,440 about maximum crowding and maximum expectation 844 00:52:35,440 --> 00:52:40,550 of crowding that would dictate a minimum frequency 845 00:52:40,550 --> 00:52:41,740 for the service. 846 00:52:41,740 --> 00:52:46,730 For example, for Green line, the maximum passengers per car 847 00:52:46,730 --> 00:52:51,950 should be no more than 225% of the seats in the peak period. 848 00:52:51,950 --> 00:52:54,110 In the off peak, the maximum passengers 849 00:52:54,110 --> 00:52:56,450 should be no more than the seated capacity, 850 00:52:56,450 --> 00:53:00,610 except in the central subway, which can go up to 140%. 851 00:53:03,610 --> 00:53:09,490 Yeah, so here is the TransLink guideline 852 00:53:09,490 --> 00:53:11,500 on frequency objective. 853 00:53:11,500 --> 00:53:17,080 It basically defines different headways based on time of days 854 00:53:17,080 --> 00:53:22,600 and based on modes of transit that the transit agencies 855 00:53:22,600 --> 00:53:26,005 and operators of the system need to satisfy. 856 00:53:28,900 --> 00:53:35,850 So in terms of guaranteeing that the load and the crowding 857 00:53:35,850 --> 00:53:38,940 situation is addressed in the guideline that 858 00:53:38,940 --> 00:53:44,070 is related to frequency, there is some analysis 859 00:53:44,070 --> 00:53:51,180 that is done to make sure that your frequency is high enough 860 00:53:51,180 --> 00:53:54,630 that your crowding situation does not 861 00:53:54,630 --> 00:53:56,490 exceed a certain level. 862 00:53:56,490 --> 00:54:00,480 Why do we care about the loads and the crowding situation? 863 00:54:00,480 --> 00:54:02,820 The first reason is that we do not want 864 00:54:02,820 --> 00:54:07,740 to frequently deny to board. 865 00:54:07,740 --> 00:54:10,830 That would yield to increased travel time. 866 00:54:10,830 --> 00:54:16,170 And we would like to basically avoid higher dwell time that is 867 00:54:16,170 --> 00:54:19,620 associated with higher number of boardings-- 868 00:54:19,620 --> 00:54:23,070 alighting number of wheelchairs on board-- 869 00:54:23,070 --> 00:54:28,800 that would eventually, again, increase the running time. 870 00:54:28,800 --> 00:54:31,160 For example, in the acceptable-- 871 00:54:31,160 --> 00:54:36,780 in a regular standard 40-foot bus 872 00:54:36,780 --> 00:54:40,960 the acceptable load of 70 passengers are considered. 873 00:54:40,960 --> 00:54:45,330 However, the average that we estimate based on our design 874 00:54:45,330 --> 00:54:48,300 should not exceed 55 passengers. 875 00:54:48,300 --> 00:54:50,640 So that's the case of indeterminacy 876 00:54:50,640 --> 00:54:55,770 and the confidence intervals that we kind of allow 877 00:54:55,770 --> 00:54:58,300 in the analysis. 878 00:54:58,300 --> 00:55:00,750 However, it's a little different with the equity issue 879 00:55:00,750 --> 00:55:05,700 that you mentioned in this case because you want to make sure 880 00:55:05,700 --> 00:55:09,810 that to a certain degree of confidence 881 00:55:09,810 --> 00:55:15,130 your system is performing at an acceptable crowding level. 882 00:55:15,130 --> 00:55:17,760 The variability to demand could actually 883 00:55:17,760 --> 00:55:22,030 relate to bulk arrivals or group arrivals 884 00:55:22,030 --> 00:55:27,450 that may happen at once due to exchanges 885 00:55:27,450 --> 00:55:30,120 from different service. 886 00:55:30,120 --> 00:55:35,280 Or it could actually relate to the variability in the headway 887 00:55:35,280 --> 00:55:39,720 or variability in the nature of demand based on day of week, 888 00:55:39,720 --> 00:55:43,470 based on seasonality, based on the weather condition, 889 00:55:43,470 --> 00:55:45,810 and so on and so forth. 890 00:55:45,810 --> 00:55:52,640 Then the other objective in the guideline of TransLink 891 00:55:52,640 --> 00:55:57,020 is the objective of comfort. 892 00:55:57,020 --> 00:56:00,440 And it basically deals, again, with the crowding level 893 00:56:00,440 --> 00:56:04,040 or the design crowding level or crowding averages 894 00:56:04,040 --> 00:56:09,380 that we are basically computing as objectives to be 895 00:56:09,380 --> 00:56:11,330 tested for this guideline. 896 00:56:11,330 --> 00:56:14,390 As you can see again, for different peak 897 00:56:14,390 --> 00:56:20,950 periods and different type of vehicles and type of service, 898 00:56:20,950 --> 00:56:27,530 there is published numbers in terms of the averages 899 00:56:27,530 --> 00:56:36,810 or computed averages for the maximum crowding levels. 900 00:56:36,810 --> 00:56:40,620 So look at these numbers. 901 00:56:40,620 --> 00:56:45,510 Does anyone have any idea of why some of these numbers 902 00:56:45,510 --> 00:56:46,780 are larger? 903 00:56:46,780 --> 00:56:51,570 For example, for peak 15 minutes in AM and PM peak, 904 00:56:51,570 --> 00:56:57,600 you would allow an average of 60 passengers on-board. 905 00:56:57,600 --> 00:57:00,780 However, when it comes to computations of peak 906 00:57:00,780 --> 00:57:05,327 30 minutes, you would only allow 55. 907 00:57:05,327 --> 00:57:07,701 AUDIENCE: You don't want people to be left behind because 908 00:57:07,701 --> 00:57:09,027 of frequency of the services. 909 00:57:09,027 --> 00:57:13,220 Knowing that there maybe another hour or service 910 00:57:13,220 --> 00:57:14,456 is another five minutes. 911 00:57:14,456 --> 00:57:16,440 NEEMA NASSIR: Right, yeah, that could 912 00:57:16,440 --> 00:57:20,340 be one reason that actually yielded to this design. 913 00:57:20,340 --> 00:57:24,454 Any other thoughts or suggestions? 914 00:57:24,454 --> 00:57:27,490 AUDIENCE: Yeah, the 15 minute peaking limit 915 00:57:27,490 --> 00:57:29,407 allows you to have one very full vehicle that 916 00:57:29,407 --> 00:57:31,198 sort of would skew the averages [INAUDIBLE] 917 00:57:31,198 --> 00:57:33,929 and could moderate that. 918 00:57:33,929 --> 00:57:34,720 NEEMA NASSIR: Yeah. 919 00:57:34,720 --> 00:57:35,594 AUDIENCE: [INAUDIBLE] 920 00:57:35,594 --> 00:57:38,069 NEEMA NASSIR: Yeah, exactly. 921 00:57:38,069 --> 00:57:38,610 Yeah, please. 922 00:57:38,610 --> 00:57:40,011 AUDIENCE: Is this like-- 923 00:57:40,011 --> 00:57:42,610 could be, like, a chicken or egg problem 924 00:57:42,610 --> 00:57:47,020 where the ones the routes were needed with peak 15 minute 925 00:57:47,020 --> 00:57:50,585 headway are ones that you will expect a lot of people. 926 00:57:50,585 --> 00:57:53,085 Or is it the other way around where you see a lot of people, 927 00:57:53,085 --> 00:57:55,810 and then you make it [INAUDIBLE].. 928 00:57:55,810 --> 00:57:57,160 NEEMA NASSIR: Yes, right. 929 00:57:57,160 --> 00:58:00,110 Well, yeah, that's exactly one of the considerations 930 00:58:00,110 --> 00:58:05,590 in design of a frequency and design of a schedule. 931 00:58:05,590 --> 00:58:10,810 So if I want to reward-- 932 00:58:10,810 --> 00:58:14,380 what Ari mentioned-- if you're computing your average over 15 933 00:58:14,380 --> 00:58:18,910 minutes and then you're computing your average over 13 934 00:58:18,910 --> 00:58:21,950 minutes and you're observing a higher crowding level over 935 00:58:21,950 --> 00:58:30,730 the accepted level, then if the situation becomes more 936 00:58:30,730 --> 00:58:32,620 critical-- 937 00:58:32,620 --> 00:58:36,526 if you're computing over 30 minutes, 938 00:58:36,526 --> 00:58:37,900 it's more critical than if you're 939 00:58:37,900 --> 00:58:41,000 computing over 15 minutes. 940 00:58:41,000 --> 00:58:45,940 So if you observe an average of high crowding 941 00:58:45,940 --> 00:58:49,750 level for 30 minutes or for a longer period, 942 00:58:49,750 --> 00:58:53,560 that is a sign of a higher actual average 943 00:58:53,560 --> 00:58:57,711 of crowding level. 944 00:58:57,711 --> 00:58:58,210 Yeah. 945 00:58:58,210 --> 00:58:58,630 AUDIENCE: It's funny. 946 00:58:58,630 --> 00:59:00,850 NEEMA NASSIR: Yeah, maybe you explained it better. 947 00:59:00,850 --> 00:59:02,099 I shouldn't have rewarded you. 948 00:59:04,640 --> 00:59:08,420 So the other objective that TransLink covers 949 00:59:08,420 --> 00:59:12,410 is basically the comfort situation. 950 00:59:12,410 --> 00:59:14,360 This is one of the interesting ones. 951 00:59:14,360 --> 00:59:15,980 And I do not really know how it's 952 00:59:15,980 --> 00:59:18,230 possible to measure this for a design. 953 00:59:18,230 --> 00:59:23,960 It's probably usually being used in evaluation of the existing 954 00:59:23,960 --> 00:59:24,920 service. 955 00:59:24,920 --> 00:59:30,020 It basically deals with number of minutes 956 00:59:30,020 --> 00:59:38,480 that you expect from passengers to be standing-- 957 00:59:38,480 --> 00:59:42,320 maximum number of minutes of being standing on board. 958 00:59:42,320 --> 00:59:45,500 And there's a level of compliance 959 00:59:45,500 --> 00:59:47,250 that is defined for each one. 960 00:59:47,250 --> 00:59:53,840 For example, you would allow no more than 95% cases 961 00:59:53,840 --> 01:00:00,140 that passengers would have to stand more than 20 minutes 962 01:00:00,140 --> 01:00:01,940 on the trip. 963 01:00:01,940 --> 01:00:05,510 So I can imagine if you have an existing system 964 01:00:05,510 --> 01:00:09,170 and if you want to measure this, you 965 01:00:09,170 --> 01:00:13,220 can go out and survey these numbers 966 01:00:13,220 --> 01:00:18,090 and compute averages and compute the distributions. 967 01:00:18,090 --> 01:00:24,440 However, if you have a design which is not in place, 968 01:00:24,440 --> 01:00:25,940 does anyone have any thoughts how 969 01:00:25,940 --> 01:00:33,406 you can come up with some sort of estimation for this case? 970 01:00:33,406 --> 01:00:35,354 AUDIENCE: [INAUDIBLE] 971 01:00:35,354 --> 01:00:37,760 NEEMA NASSIR: Yeah, simulation could be 972 01:00:37,760 --> 01:00:39,970 one possible way of doing this. 973 01:00:39,970 --> 01:00:48,566 Simulation of demand could be one possible way of doing this. 974 01:00:48,566 --> 01:00:52,570 AUDIENCE: You could check, like, if it's a train-- 975 01:00:52,570 --> 01:00:57,850 I saw the Sky Train there, you can see where the train fills 976 01:00:57,850 --> 01:01:00,280 up and then assume that everyone else-- 977 01:01:00,280 --> 01:01:02,510 everyone that boards the train afterwards-- 978 01:01:02,510 --> 01:01:05,147 if you know how many people are boarding-- if you have-- 979 01:01:05,147 --> 01:01:05,980 NEEMA NASSIR: Right. 980 01:01:05,980 --> 01:01:07,720 AUDIENCE: --data to do that, then you 981 01:01:07,720 --> 01:01:11,050 can assume that everyone who's boarding after is standing. 982 01:01:11,050 --> 01:01:15,220 And then you can see how many of them have 20 minutes or more 983 01:01:15,220 --> 01:01:17,854 left on the journey till the train [INAUDIBLE] 984 01:01:17,854 --> 01:01:18,770 NEEMA NASSIR: Exactly. 985 01:01:18,770 --> 01:01:19,600 That's a good idea. 986 01:01:19,600 --> 01:01:24,220 So what you're suggesting is to generate the load profile 987 01:01:24,220 --> 01:01:26,500 for the line. 988 01:01:26,500 --> 01:01:30,610 And whenever the load is over the seating capacity, 989 01:01:30,610 --> 01:01:33,170 you measure the time period on that. 990 01:01:33,170 --> 01:01:35,150 So that's a good idea. 991 01:01:35,150 --> 01:01:36,280 Right. 992 01:01:36,280 --> 01:01:41,170 So in terms of for Toronto Transit Commission, 993 01:01:41,170 --> 01:01:44,320 the loading standards is something like this. 994 01:01:44,320 --> 01:01:46,330 For different types of vehicle, you 995 01:01:46,330 --> 01:01:50,840 would allow different averages for your design. 996 01:01:50,840 --> 01:01:55,120 So there is also one interesting point here. 997 01:01:55,120 --> 01:01:58,330 Why do you think for low frequency service 998 01:01:58,330 --> 01:02:06,230 you would allow a lower crowding level or a lower average 999 01:02:06,230 --> 01:02:06,969 ridership? 1000 01:02:10,961 --> 01:02:12,957 AUDIENCE: If it was that crowded, 1001 01:02:12,957 --> 01:02:16,217 you would want to add frequency to better serve people riding. 1002 01:02:16,217 --> 01:02:17,300 NEEMA NASSIR: Right, yeah. 1003 01:02:17,300 --> 01:02:22,670 That could be one good answer. 1004 01:02:22,670 --> 01:02:24,740 The other reason that I can imagine 1005 01:02:24,740 --> 01:02:27,530 is that if it's a low frequency service 1006 01:02:27,530 --> 01:02:30,110 and if someone is denied boarding to a low frequency 1007 01:02:30,110 --> 01:02:32,780 service, they have to wait longer. 1008 01:02:32,780 --> 01:02:37,700 So we would allow a little bit of a slackness in the design, 1009 01:02:37,700 --> 01:02:39,530 so we would avoid a situation like that. 1010 01:02:48,390 --> 01:02:50,560 OK, so let's stop here. 1011 01:02:50,560 --> 01:02:52,720 And then we will continue through the rest 1012 01:02:52,720 --> 01:02:59,620 of the guidelines, particularly with reliability of service 1013 01:02:59,620 --> 01:03:00,220 next week. 1014 01:03:00,220 --> 01:03:01,803 AUDIENCE: So there's a question there? 1015 01:03:01,803 --> 01:03:02,685 NEEMA NASSIR: Right. 1016 01:03:02,685 --> 01:03:03,999 AUDIENCE: I was going to ask is there a reason why we 1017 01:03:03,999 --> 01:03:05,150 use so many Canadian transits? 1018 01:03:05,150 --> 01:03:05,990 NEEMA NASSIR: Sorry? 1019 01:03:05,990 --> 01:03:07,406 AUDIENCE: Is there a reason why we 1020 01:03:07,406 --> 01:03:09,510 use so many Canadian transits? 1021 01:03:09,510 --> 01:03:11,335 NEEMA NASSIR: That's a good question. 1022 01:03:11,335 --> 01:03:13,400 AUDIENCE: Their public transportation is better. 1023 01:03:13,400 --> 01:03:15,430 NEEMA NASSIR: Yeah, better-- probably the closest place 1024 01:03:15,430 --> 01:03:16,720 that has better public transportation. 1025 01:03:16,720 --> 01:03:17,220 [LAUGHTER] 1026 01:03:17,220 --> 01:03:20,670 AUDIENCE: [LAUGHTER]