WEBVTT 1 00:00:00.000 --> 00:00:02.490 Hi, and welcome to module four 2 00:00:02.490 --> 00:00:03.880 of Advanced GIS. 3 00:00:03.880 --> 00:00:06.340 This is the first of two modules we'll spend focused 4 00:00:06.340 --> 00:00:09.800 on moving things around the landscape using GIS. 5 00:00:09.800 --> 00:00:11.540 In this module, it's people. 6 00:00:11.540 --> 00:00:13.503 Next week, it's water. 7 00:00:14.380 --> 00:00:16.650 I thought I'd change things up a bit this week 8 00:00:16.650 --> 00:00:18.840 by starting with a story about bridges. 9 00:00:18.840 --> 00:00:21.180 Really old bridges, in fact. 10 00:00:21.180 --> 00:00:23.620 And while we're starting with a story about bridges, 11 00:00:23.620 --> 00:00:25.420 we'll end with an approach to network analysis 12 00:00:25.420 --> 00:00:27.900 that relies heavily on a branch of theoretical mathematics 13 00:00:27.900 --> 00:00:29.823 known as graph theory. 14 00:00:30.720 --> 00:00:32.740 We'll explore how to apply network analysis techniques 15 00:00:32.740 --> 00:00:35.103 to solve transportation problems. 16 00:00:36.270 --> 00:00:37.600 Graph theory is a branch of mathematics 17 00:00:37.600 --> 00:00:40.500 that uses structures or graphs 18 00:00:40.500 --> 00:00:42.440 to model pairwise relationships. 19 00:00:42.440 --> 00:00:44.500 In this case, we're interested in modeling 20 00:00:44.500 --> 00:00:46.980 the spatial relationship between discrete 21 00:00:46.980 --> 00:00:49.140 locations on a map. 22 00:00:49.140 --> 00:00:52.060 So what does that have to do with the image on the right? 23 00:00:52.060 --> 00:00:54.620 Well, this is where it gets interesting. 24 00:00:54.620 --> 00:00:58.770 Back in the 1730s, Leonard Euler was wondering 25 00:00:58.770 --> 00:01:00.820 about the seven bridges of Konigsberg. 26 00:01:00.820 --> 00:01:03.300 Specifically, he was wondering if he could cross 27 00:01:03.300 --> 00:01:06.880 all seven bridges one time without having to travel 28 00:01:06.880 --> 00:01:09.060 across the same bridge twice. 29 00:01:09.060 --> 00:01:12.340 He correctly concluded that the answer was no, 30 00:01:12.340 --> 00:01:14.600 but never proved as much. 31 00:01:14.600 --> 00:01:18.540 More than 100 years later, Carl Hierholzer came up 32 00:01:18.540 --> 00:01:21.140 with the proof known as Euler's Theorem. 33 00:01:21.140 --> 00:01:23.973 This eventually laid the groundwork for graph theory. 34 00:01:26.640 --> 00:01:28.320 If we reduce each landmass to a point 35 00:01:28.320 --> 00:01:30.760 and convert each bridge into a line 36 00:01:30.760 --> 00:01:33.440 representing the connections between the landmasses, 37 00:01:33.440 --> 00:01:36.600 we are able to abstract a physical landscape 38 00:01:36.600 --> 00:01:40.440 into a graphical representation using points and lines. 39 00:01:40.440 --> 00:01:44.580 A graph of, for example, a road network 40 00:01:44.580 --> 00:01:45.820 lends itself readily to point 41 00:01:45.820 --> 00:01:47.960 and line vector representations, 42 00:01:47.960 --> 00:01:50.320 where the points represent junctions in the network 43 00:01:50.320 --> 00:01:52.950 and the lines represent the edges of the network 44 00:01:52.950 --> 00:01:54.320 to traverse in order to travel 45 00:01:54.320 --> 00:01:56.860 from one location to another. 46 00:01:56.860 --> 00:01:59.680 So what we see on the right should look familiar. 47 00:01:59.680 --> 00:02:03.300 These points and lines, or junctions and edges, 48 00:02:03.300 --> 00:02:05.300 comprise what is known as a network dataset, 49 00:02:05.300 --> 00:02:07.920 and we'll talk about what differentiates it 50 00:02:07.920 --> 00:02:09.260 from the more traditional vector data 51 00:02:09.260 --> 00:02:11.103 we've worked with previously. 52 00:02:12.210 --> 00:02:13.320 Graph theory provides the foundation 53 00:02:13.320 --> 00:02:17.250 for transportation analysis of many types, 54 00:02:17.250 --> 00:02:19.220 from simple paths to service area polygons 55 00:02:19.220 --> 00:02:22.170 to origin destination matrices. 56 00:02:22.170 --> 00:02:24.630 You could even identify weak links in the network, 57 00:02:24.630 --> 00:02:27.480 the loss of which would cause severe travel disruptions. 58 00:02:27.480 --> 00:02:30.280 Routing, logistics, and even site selection 59 00:02:30.280 --> 00:02:33.513 can benefit from knowledge gleaned from network analysis. 60 00:02:34.780 --> 00:02:36.320 Let's look at a few example graphs 61 00:02:36.320 --> 00:02:39.450 and consider some routing alternatives. 62 00:02:39.450 --> 00:02:41.480 We want to travel from point one to point nine 63 00:02:41.480 --> 00:02:43.400 along the available edges, 64 00:02:43.400 --> 00:02:46.540 without traversing the same edge more than once. 65 00:02:46.540 --> 00:02:49.710 There are four possible routes available to us. 66 00:02:49.710 --> 00:02:51.860 For example, we can travel from point one 67 00:02:51.860 --> 00:02:55.533 to point two to point seven to point nine. 68 00:02:57.100 --> 00:03:00.033 What if we add direction information to the edges? 69 00:03:01.300 --> 00:03:04.160 Those could be used to manage the flow of traffic. 70 00:03:04.160 --> 00:03:06.520 All the same conditions hold from before. 71 00:03:06.520 --> 00:03:09.510 We still want to travel from point one to point nine, 72 00:03:09.510 --> 00:03:10.740 and we still don't want to traverse 73 00:03:10.740 --> 00:03:12.400 an edge more than once. 74 00:03:12.400 --> 00:03:16.040 In addition, we need to respect the direction arrows. 75 00:03:16.040 --> 00:03:20.613 As a result, we go from four possible routes down to two. 76 00:03:22.440 --> 00:03:25.040 Let's take it one step further and add a weight, 77 00:03:25.040 --> 00:03:28.600 or a cost, associated with traversing each edge. 78 00:03:28.600 --> 00:03:30.510 In the last two examples, 79 00:03:30.510 --> 00:03:33.690 it was about traveling from point one to point nine. 80 00:03:33.690 --> 00:03:35.480 Now, let's figure out the best route 81 00:03:35.480 --> 00:03:37.680 between these two locations. 82 00:03:37.680 --> 00:03:40.050 I've assigned a cost for each edge, 83 00:03:40.050 --> 00:03:41.580 and to determine the best route, 84 00:03:41.580 --> 00:03:43.640 we need to sum the cost of all edges traversed 85 00:03:43.640 --> 00:03:47.850 on the route between point one and point nine. 86 00:03:47.850 --> 00:03:50.140 We see that route one costs 17 units, 87 00:03:50.140 --> 00:03:52.893 while route two costs 13 units. 88 00:03:53.940 --> 00:03:56.260 As a result, we declare that route two, 89 00:03:56.260 --> 00:03:59.260 the path from point one to point two to point six 90 00:03:59.260 --> 00:04:01.840 to point nine, is the least cost path 91 00:04:01.840 --> 00:04:04.580 among all the available options. 92 00:04:04.580 --> 00:04:06.660 Least cost path is fancy speak 93 00:04:06.660 --> 00:04:09.630 for shortest route in this case. 94 00:04:09.630 --> 00:04:13.203 Note again that cost is not always measured in distance. 95 00:04:14.800 --> 00:04:17.370 Okay, so that was pretty easy, 96 00:04:17.370 --> 00:04:19.340 but what about something like this? 97 00:04:19.340 --> 00:04:21.690 There's no way I'm going to do all that math 98 00:04:21.690 --> 00:04:23.080 on the back of an envelope. 99 00:04:23.080 --> 00:04:25.170 It's time to turn the heavy lifting 100 00:04:25.170 --> 00:04:26.583 over to the computers. 101 00:04:27.930 --> 00:04:30.360 Remember that seven bridges problem? 102 00:04:30.360 --> 00:04:33.210 Well, fast forward 75 years, 103 00:04:33.210 --> 00:04:35.060 and Edsger Dijkstra developed an algorithm 104 00:04:35.060 --> 00:04:38.560 which computed the shortest path between nodes in a graph. 105 00:04:38.560 --> 00:04:40.600 This discovery had significant implications 106 00:04:40.600 --> 00:04:43.100 for any number of routing applications, 107 00:04:43.100 --> 00:04:47.220 including transportation modeling, utilities analysis, 108 00:04:47.220 --> 00:04:50.133 and computer science among other topics. 109 00:04:51.520 --> 00:04:56.133 Now let's look at the network analysis toolkit in ArcGIS. 110 00:04:57.280 --> 00:05:00.200 Just a quick diversion though before we get started. 111 00:05:00.200 --> 00:05:02.560 I mentioned earlier that a network data set 112 00:05:02.560 --> 00:05:04.770 is different than the vector data 113 00:05:04.770 --> 00:05:06.380 we've worked with previously. 114 00:05:06.380 --> 00:05:08.160 Think of a network data set 115 00:05:08.160 --> 00:05:09.900 as a special kind of vector data set. 116 00:05:09.900 --> 00:05:12.270 Special because there's quite a bit 117 00:05:12.270 --> 00:05:14.740 of additional information baked into the network data set 118 00:05:14.740 --> 00:05:16.833 to enable analysis. 119 00:05:17.670 --> 00:05:18.840 We'll talk about this in the context 120 00:05:18.840 --> 00:05:20.580 of the road network, 121 00:05:20.580 --> 00:05:23.100 but all the same concepts would apply to sidewalks, 122 00:05:23.100 --> 00:05:26.360 hiking trails, bike lanes, and so on. 123 00:05:26.360 --> 00:05:29.480 So the network data set includes the roads, 124 00:05:29.480 --> 00:05:32.020 the intersections or junctions, 125 00:05:32.020 --> 00:05:34.340 and some notion of connectivity. 126 00:05:34.340 --> 00:05:38.490 The roads are straightforward, intersections too, 127 00:05:38.490 --> 00:05:39.940 but what about connectivity? 128 00:05:39.940 --> 00:05:42.480 The easiest example here is to think 129 00:05:42.480 --> 00:05:43.480 of an interstate overpass 130 00:05:43.480 --> 00:05:45.720 and the road that runs underneath it. 131 00:05:45.720 --> 00:05:47.640 You can't turn from one to the other 132 00:05:47.640 --> 00:05:49.520 unless there's a ramp. 133 00:05:49.520 --> 00:05:51.720 There's a vertical offset that restricts 134 00:05:51.720 --> 00:05:53.880 the connectivity of the two roads. 135 00:05:53.880 --> 00:05:56.180 You can also apply specific restrictions 136 00:05:56.180 --> 00:05:58.740 like one-way streets or turn limits 137 00:05:58.740 --> 00:06:00.760 at particular intersections. 138 00:06:00.760 --> 00:06:03.480 Maybe you're not allowed to turn right or left. 139 00:06:03.480 --> 00:06:06.260 That affects everything from individual route decisions 140 00:06:06.260 --> 00:06:09.040 to cumulative traffic flows. 141 00:06:09.040 --> 00:06:11.080 The last piece of information embedded 142 00:06:11.080 --> 00:06:13.820 in the network data set is the travel cost. 143 00:06:13.820 --> 00:06:17.120 We typically think of travel cost and time units, 144 00:06:17.120 --> 00:06:19.040 but we could also consider fuel consumption 145 00:06:19.040 --> 00:06:22.503 or calories or total distance instead. 146 00:06:23.520 --> 00:06:25.960 Calculating the time with no congestion is easy. 147 00:06:25.960 --> 00:06:29.190 Divide the length of the segment by the speed limit. 148 00:06:29.190 --> 00:06:30.760 Total distance isn't hard either, 149 00:06:30.760 --> 00:06:33.200 but if you're interested in other cost alternatives, 150 00:06:33.200 --> 00:06:35.820 you may need to develop your own cost model 151 00:06:35.820 --> 00:06:39.120 to calculate those values based on other factors, 152 00:06:39.120 --> 00:06:42.200 just like you would for time or distance. 153 00:06:42.200 --> 00:06:44.100 One last thing before we move on 154 00:06:44.100 --> 00:06:45.820 to the network analyst toolkit. 155 00:06:45.820 --> 00:06:49.320 A network analysis requires a network data set. 156 00:06:49.320 --> 00:06:52.500 This data set can be hosted either locally, 157 00:06:52.500 --> 00:06:55.140 a data set on your local computer or network, 158 00:06:55.140 --> 00:06:58.740 or can be accessed through cloud-based resources. 159 00:06:58.740 --> 00:07:01.340 One thing to note here is that ArcGIS credits 160 00:07:01.340 --> 00:07:04.520 are consumed if you use the cloud-based approach. 161 00:07:04.520 --> 00:07:07.020 We don't assign you many credits for this course, 162 00:07:07.020 --> 00:07:09.520 and those that we do are used for publishing data, 163 00:07:09.520 --> 00:07:11.840 web maps, and web apps. 164 00:07:11.840 --> 00:07:13.680 If you want to perform a network analysis 165 00:07:13.680 --> 00:07:16.120 somewhere outside of Chittenden County, Vermont, 166 00:07:16.120 --> 00:07:18.390 you will need to develop a network data set 167 00:07:18.390 --> 00:07:20.313 of your own for that region. 168 00:07:21.700 --> 00:07:24.160 A network analysis workflow typically follows 169 00:07:24.160 --> 00:07:26.740 a minimum three-step procedure. 170 00:07:26.740 --> 00:07:29.373 First, create an analysis layer. 171 00:07:31.590 --> 00:07:32.580 We'll talk about the different types 172 00:07:32.580 --> 00:07:34.280 of analysis you can perform. 173 00:07:34.280 --> 00:07:37.240 Each one requires a slightly different version 174 00:07:37.240 --> 00:07:39.990 of an analysis layer. 175 00:07:39.990 --> 00:07:42.400 Once you select the type of analysis to perform, 176 00:07:42.400 --> 00:07:44.480 in this case, we'll compute the shortest path 177 00:07:44.480 --> 00:07:48.450 between two points, ArcGIS creates a feature data set 178 00:07:48.450 --> 00:07:52.113 to contain all the inputs and outputs of the analysis. 179 00:07:53.520 --> 00:07:55.400 Next, you'll add analysis objects 180 00:07:55.400 --> 00:07:58.367 to the feature data set you just created. 181 00:07:58.367 --> 00:08:00.500 These are things like stops or facilities 182 00:08:00.500 --> 00:08:04.540 or incidents or origins and destinations. 183 00:08:04.540 --> 00:08:08.900 You could also include point, line, or polygon barriers, 184 00:08:08.900 --> 00:08:12.020 network elements that restrict the flow of travel. 185 00:08:12.020 --> 00:08:13.560 The image on the right illustrates 186 00:08:13.560 --> 00:08:16.140 four different types of network analysis. 187 00:08:16.140 --> 00:08:17.940 Note that each requires slightly 188 00:08:17.940 --> 00:08:19.980 different input information. 189 00:08:19.980 --> 00:08:22.160 We'll look at examples of each in the coming slides 190 00:08:22.160 --> 00:08:25.293 and in lecture two of this learning module. 191 00:08:27.270 --> 00:08:29.560 Then, once you're happy with the parameters, 192 00:08:29.560 --> 00:08:31.940 all that's left is to click run. 193 00:08:31.940 --> 00:08:34.683 This solves for the equation you just built. 194 00:08:36.030 --> 00:08:38.340 ArcGIS will lock the features of the analysis layer 195 00:08:38.340 --> 00:08:41.440 and write the output information to the data set. 196 00:08:41.440 --> 00:08:43.820 Depending on the kind of analysis you're doing, 197 00:08:43.820 --> 00:08:45.700 the output will be represented 198 00:08:45.700 --> 00:08:49.060 as a line, polygon, or matrix. 199 00:08:49.060 --> 00:08:52.880 We'll look at examples of these in a few slides. 200 00:08:52.880 --> 00:08:54.840 Well, now what? 201 00:08:54.840 --> 00:08:56.440 That's up to you. 202 00:08:56.440 --> 00:08:58.950 Maybe that's the end point of your analysis. 203 00:08:58.950 --> 00:09:00.600 You just wanted to know the travel cost 204 00:09:00.600 --> 00:09:01.900 from point A to point B. 205 00:09:01.900 --> 00:09:03.580 If that's the case, 206 00:09:03.580 --> 00:09:06.420 open the attribute table and find the answer. 207 00:09:06.420 --> 00:09:08.780 Or maybe this is just the first step 208 00:09:08.780 --> 00:09:12.080 in a much longer geoprocessing workflow. 209 00:09:12.080 --> 00:09:14.900 The only limit here, as I've said before, 210 00:09:14.900 --> 00:09:16.473 is your imagination. 211 00:09:17.940 --> 00:09:19.650 I'd like to round out the lecture 212 00:09:19.650 --> 00:09:21.920 with a brief review of four types of network analysis 213 00:09:21.920 --> 00:09:24.520 you can perform with ArcGIS. 214 00:09:24.520 --> 00:09:26.340 We'll start with the easiest, 215 00:09:26.340 --> 00:09:29.440 but most fundamental of all network analysis operations. 216 00:09:29.440 --> 00:09:32.840 This is the algorithm that makes all the others possible. 217 00:09:32.840 --> 00:09:35.200 Let's compute a route between two point locations 218 00:09:35.200 --> 00:09:37.860 based on a specified travel cost. 219 00:09:37.860 --> 00:09:41.020 In this case, time measured in minutes. 220 00:09:41.020 --> 00:09:43.230 Remember, this information is baked into 221 00:09:43.230 --> 00:09:44.063 the network data set 222 00:09:44.063 --> 00:09:46.170 and is used by ArcGIS to determine 223 00:09:46.170 --> 00:09:48.760 the most efficient route between two points. 224 00:09:48.760 --> 00:09:51.880 On the right, we see the output from the analysis. 225 00:09:51.880 --> 00:09:54.963 It's a line connecting my two point locations. 226 00:09:55.800 --> 00:09:57.440 If we open the attribute table 227 00:09:57.440 --> 00:09:59.200 for the routes feature class, 228 00:09:59.200 --> 00:10:00.740 we see that there is a single record 229 00:10:00.740 --> 00:10:03.920 defining my travel between stops one and two. 230 00:10:03.920 --> 00:10:06.380 The record includes information on the total distance 231 00:10:06.380 --> 00:10:07.960 measured in meters, 232 00:10:07.960 --> 00:10:10.620 and the total cost measured in minutes. 233 00:10:10.620 --> 00:10:12.740 Remember, both of those unit definitions 234 00:10:12.740 --> 00:10:15.543 can be found in the network data set properties. 235 00:10:17.740 --> 00:10:20.490 Let's up the difficulty just a bit. 236 00:10:20.490 --> 00:10:23.460 In this case, we'll perform a closest facility analysis. 237 00:10:23.460 --> 00:10:25.040 This type of analysis identifies 238 00:10:25.040 --> 00:10:27.940 the closest facility to an incident. 239 00:10:27.940 --> 00:10:29.580 Note that it's possible to have 240 00:10:29.580 --> 00:10:30.960 both more than one incident 241 00:10:30.960 --> 00:10:34.120 and more than one closest facility identified. 242 00:10:34.120 --> 00:10:36.510 We see the simplest case on the right, 243 00:10:36.510 --> 00:10:38.180 where I've identified the single closest facility 244 00:10:38.180 --> 00:10:40.120 to a single incident. 245 00:10:40.120 --> 00:10:42.920 Consider you're an emergency services operator. 246 00:10:42.920 --> 00:10:44.260 There's a fire. 247 00:10:44.260 --> 00:10:45.900 You would like to know which fire station 248 00:10:45.900 --> 00:10:47.620 is closest to the incident. 249 00:10:47.620 --> 00:10:49.760 We see that the route delineated 250 00:10:49.760 --> 00:10:52.110 as a purple line connecting a circle, 251 00:10:52.110 --> 00:10:53.600 the location of a fire station, 252 00:10:53.600 --> 00:10:56.313 with a square, the location of the fire. 253 00:10:57.780 --> 00:11:00.320 That information is written to the routes feature class 254 00:11:00.320 --> 00:11:03.340 within the closest facility feature data set. 255 00:11:03.340 --> 00:11:06.020 If I open my route attribute table, 256 00:11:06.020 --> 00:11:07.740 I once again see a single record 257 00:11:07.740 --> 00:11:10.980 describing travel between locations one and three. 258 00:11:10.980 --> 00:11:14.220 The total distance is just under 4,800 meters, 259 00:11:14.220 --> 00:11:15.700 and the total travel time 260 00:11:15.700 --> 00:11:19.503 from the closest fire station is just under six minutes. 261 00:11:22.380 --> 00:11:23.580 Next up is what's known 262 00:11:23.580 --> 00:11:26.610 as an origin destination cost matrix. 263 00:11:26.610 --> 00:11:28.440 In this case, we're computing the travel time 264 00:11:28.440 --> 00:11:30.280 between every origin and every destination 265 00:11:30.280 --> 00:11:32.700 in our data sets. 266 00:11:32.700 --> 00:11:35.180 This information is written out as a cost matrix. 267 00:11:35.180 --> 00:11:37.460 The image on the right illustrates the connections 268 00:11:37.460 --> 00:11:40.623 from each origin to each destination. 269 00:11:41.780 --> 00:11:43.980 If we open the lines attribute table, 270 00:11:43.980 --> 00:11:47.010 we see the details of each route. 271 00:11:47.010 --> 00:11:48.400 This information could be used to identify 272 00:11:48.400 --> 00:11:50.980 the closest origin to each destination, 273 00:11:50.980 --> 00:11:54.780 or more frequently is used as a primary input data set 274 00:11:54.780 --> 00:11:57.560 into more formalized travel demand models. 275 00:11:57.560 --> 00:12:00.080 One other thing about the attribute table. 276 00:12:00.080 --> 00:12:01.940 Notice that the total meters attribute 277 00:12:01.940 --> 00:12:03.820 is full of null values. 278 00:12:03.820 --> 00:12:07.500 In this case, only the travel cost is measured. 279 00:12:07.500 --> 00:12:08.500 Although the shape length attribute 280 00:12:08.500 --> 00:12:12.380 in the previous examples match the total meters attribute, 281 00:12:12.380 --> 00:12:14.420 in this case, the shape length refers 282 00:12:14.420 --> 00:12:16.760 to the length of the straight line. 283 00:12:16.760 --> 00:12:19.400 Not really useful for the level of detail you'd need 284 00:12:19.400 --> 00:12:21.600 for network analysis. 285 00:12:21.600 --> 00:12:23.840 Otherwise, you could just use the near tool 286 00:12:23.840 --> 00:12:26.793 if you weren't interested in travel across a network. 287 00:12:27.780 --> 00:12:31.520 The last of the examples looks at a service area analysis. 288 00:12:31.520 --> 00:12:33.940 This type of analysis defines polygon boundaries 289 00:12:33.940 --> 00:12:36.260 that represent the area you could travel to 290 00:12:36.260 --> 00:12:37.720 within a given amount of time 291 00:12:37.720 --> 00:12:40.520 from a given point of origin. 292 00:12:40.520 --> 00:12:42.700 In the example here, we've got two facilities, 293 00:12:42.700 --> 00:12:45.580 the purple dots and polygons representing 294 00:12:45.580 --> 00:12:46.700 the area of travel 295 00:12:46.700 --> 00:12:49.660 within one and three minute cutoff times. 296 00:12:49.660 --> 00:12:51.920 If we look at the attribute table, 297 00:12:51.920 --> 00:12:54.020 we can see the facility ID, 298 00:12:54.020 --> 00:12:56.840 the from and to breaks representing the cutoff times 299 00:12:56.840 --> 00:12:59.680 you specify, and the typical system fields 300 00:12:59.680 --> 00:13:02.100 like shape length and shape area. 301 00:13:02.100 --> 00:13:03.740 There are four records, 302 00:13:03.740 --> 00:13:05.840 two polygons for each facility representing 303 00:13:05.840 --> 00:13:09.200 their one and three minute cutoff boundaries. 304 00:13:09.200 --> 00:13:13.080 Service areas are useful for locating gaps in coverage. 305 00:13:13.080 --> 00:13:16.800 Consider our first responder example from before. 306 00:13:16.800 --> 00:13:19.460 Want to locate service gaps or gaps in coverage? 307 00:13:19.460 --> 00:13:23.370 A service area analysis would be a good place to start. 308 00:13:23.370 --> 00:13:24.560 Maybe you want to open a gas station 309 00:13:24.560 --> 00:13:26.540 or a coffee shop or a hotel. 310 00:13:26.540 --> 00:13:29.760 A service area could identify priority locations 311 00:13:29.760 --> 00:13:30.700 if you want to be close to 312 00:13:30.700 --> 00:13:33.483 or far away from your competitors. 313 00:13:35.160 --> 00:13:36.280 Two other analyses I'll mention 314 00:13:36.280 --> 00:13:37.620 without covering in any depth 315 00:13:37.620 --> 00:13:39.400 are location allocation analysis 316 00:13:39.400 --> 00:13:41.700 and vehicle routing problems. 317 00:13:41.700 --> 00:13:43.770 These are both good skills for working 318 00:13:43.770 --> 00:13:45.780 with vehicle fleets, delivery scheduling, 319 00:13:45.780 --> 00:13:47.613 and the transport of goods. 320 00:13:48.680 --> 00:13:50.820 Two more elements of note though, 321 00:13:50.820 --> 00:13:52.040 before I offer a short demonstration 322 00:13:52.040 --> 00:13:55.323 of the Network Analysis Toolkit. 323 00:13:56.400 --> 00:13:58.980 The first of these is that you can compute directions 324 00:13:58.980 --> 00:14:01.710 for route, closest facility, 325 00:14:01.710 --> 00:14:03.400 and vehicle routing analyses. 326 00:14:03.400 --> 00:14:08.400 Simply specify the output type as XML, HTML, or text. 327 00:14:09.000 --> 00:14:10.980 Determine whether to include travel cost, 328 00:14:10.980 --> 00:14:13.720 and if so, specify the time attribute. 329 00:14:13.720 --> 00:14:15.660 That, of course, comes directly 330 00:14:15.660 --> 00:14:18.440 from the network dataset specification. 331 00:14:18.440 --> 00:14:20.610 Expect output here that is similar 332 00:14:20.610 --> 00:14:22.560 to what you would see from Google Maps. 333 00:14:23.420 --> 00:14:24.980 The second tool I want to mention 334 00:14:24.980 --> 00:14:27.980 is the copy traversed source features. 335 00:14:27.980 --> 00:14:30.100 This is a great way to export the results 336 00:14:30.100 --> 00:14:33.000 of a particular network analysis. 337 00:14:33.000 --> 00:14:34.780 The network analysis layer naming scheme 338 00:14:34.780 --> 00:14:38.100 in the geodatabase is a bit difficult to parse. 339 00:14:38.100 --> 00:14:40.140 Use this tool to export your results 340 00:14:40.140 --> 00:14:43.140 while having control over the naming conventions. 341 00:14:43.140 --> 00:14:46.830 The output of this operation is two feature classes, 342 00:14:46.830 --> 00:14:50.883 the edges and the junctions, and a table for turns. 343 00:14:51.720 --> 00:14:54.360 This data offers a much more detailed documentation 344 00:14:54.360 --> 00:14:56.260 of each of the routes that were identified 345 00:14:56.260 --> 00:14:59.460 in the analysis, broken down by individual segment 346 00:14:59.460 --> 00:15:01.263 within an overall route. 347 00:15:02.400 --> 00:15:04.520 There's the route ID attribute, 348 00:15:04.520 --> 00:15:06.540 plus the from and to junctions 349 00:15:06.540 --> 00:15:07.830 for each of those segments, 350 00:15:07.830 --> 00:15:11.070 along with cost and distance calculations. 351 00:15:11.070 --> 00:15:13.110 That's a lot of information. 352 00:15:13.110 --> 00:15:14.670 Okay, I think that's enough 353 00:15:14.670 --> 00:15:16.760 to whet your network analysis appetites. 354 00:15:16.760 --> 00:15:18.630 In the next lecture, I'll run through 355 00:15:18.630 --> 00:15:21.120 a short demonstration of a few of these techniques. 356 00:15:21.120 --> 00:15:22.383 See you in a minute.