Network analysis can be used to solve many different transportation problems that would be very challenging to solve otherwise. A prerequisite to performing network analysis is that you have a network model. In ESRI’s terms, this is a network dataset. We walked through an exercise to construct a network dataset in the last lesson. Of course, it is not a requirement of network analysis that you construct your own network model. There are a number of commercially available network models you can use instead.
The types of problems which network analysis can be used to solve are quite varied. One common characteristic of the algorithms that power each is that they involve determining the cost of one or more routes through the network. The cost is most commonly based on time or distance, but you can define a cost attribute any way you want. For example, you might score each edge in the network based on its scenic value. You could then create a cost parameter based on the scenic score and use the solver to find the most scenic route.
ESRI provides 6 out-of-the-box network analyses as a part of Network Analyst. ESRI terms these network analyses “solvers.” The solvers are listed below along with a brief description of each:
The route solver determines the best route between two or more points. Most of us use this network analysis on a regular basis. Whenever you use Google Maps or a comparable service to get directions from one location to another the service is conducting a network analysis to determine the best, typically fastest, route. This solver can route any number of points according to a specified order (i.e., the traveling salesman problem) or the most efficient order.
Closest Facility Solver
This solver is used to determine the closest facility to a given location. The term "facility" can be a bit misleading. For example, this solver could be used to determine the closest ambulance to an accident scene. In this case, using ESRI’s terminology, the ambulances would be considered facilities.
Service Area Solver
The geographic region which can reach a designated facility in a certain period of time (or vice versa) is termed a service area. To determine the bounds of this area, you can use the Service Area Solver.
Origin-Destination (OD) Cost Matrix Solver
The OD cost matrix solver is generally used to determine the distances of the fastest routes between a set of origins and a set of destinations. Although the path between each origin and destination is often represented as a straight line, the route which corresponds to the time and distance costs between each pair of locations follows the street network.
Vehicle Routing Solver
The vehicle routing solver is typically used to determine the most efficient routes for a fleet of vehicles tasked with servicing a series of stops.
The location-allocation solver can be used to determine how effectively a facility site is servicing locations which have a need for its services. As such, it can be used to select the best location for a facility from a series of candidate locations.
Assignment 5-1 (15 points)
Watch this video from ESRI’s 2010 User Conference (71 minutes) which talks about conducting network analyses using the Network Analyst extension to ArcGIS. There are a number of introductory videos for network analyst available on ESRI's website some of which are more recent than the video I have referenced. If you are interested in learning more about Network Analyst, especially if you are looking to use it in ArcGIS online, I would recommend you watch some of these videos. I selected this video, since I felt it did the best job at explaining the solvers. After watching the video, address the following questions and submit your responses in the form of an M.S. Word document (no more than 500 words) to Assignment 5-1 in Canvas.
- What is the difference between a Euclidean path and a network path? (3 points)
- What is a restriction barrier? Provide an example of when you might use one. (3 points)
- What is a scaled cost barrier? Provide an example of when you might use one. (3 points)
- What type of cost attribute would you generally use to find the best driving route? How about for a walking route? (3 points)
- The service area solver and the location-allocation solver are somewhat similar. Briefly describe what differentiates them. (3 points)
Assignment 5-2 (20 points)
In Assignment 4-2 you completed Exercise 1 of ESRI’s ArcGIS Network Analyst Tutorial. For this assignment, you will complete Exercise 3 and Exercise 5. As you’re completing the exercises, address the following questions and submit your responses in the form of an M.S. Word document (no more than 500 words) to Assignment 5-2 in Canvas.
- Exercise 3: Finding the best route using a network dataset.
- Will Network Analyst use historic traffic data if you do not specify a day and time for the route? (1 points)
- Include screenshots of your fastest route before and after you implemented a point barrier. (4 points)
- Add a polygon barrier somewhere along your route and re-solve. Include a screenshot of the resulting route. (2 points)
- Find the shortest route by changing the impedance in the analysis settings to distance (i.e., meters) in the analysis settings window and re-solving. Include screenshots of both the map of the new route and the analysis settings window. (2 points)
- Did Network Analyst use historical traffic data in determining the shortest route? (1 point)
- Exercise 5: Calculating service areas and creating an OD cost matrix.
- Include a screenshot of the service area map both before and after you move a facility and re-solve the analysis. (i.e., 2 screenshots)(4 points)
- What does ArcMap do if a store is inside multiple polygons and you join the stores features to the polygon features? (2 points)
- Include a screenshot of your final map showing both the results of the OD cost analysis and the service area analysis displayed on the same map. (4 points)