Hands On Convincing the Chief Continued from page 53 3. Inspect the statistics for the sub-fourminute runs for each table, particularly the Mean and Standard Deviation values. It certainly would be nice to place this data in a table so it could be studied. To quickly post this summary data, open the Excel spreadsheet template, RunSummary1.xls, provided in the sample dataset, and copy values from the fourminute statistics generated for each station to the appropriate cells in the spreadsheet. In ArcMap, under the four-minute statistics for Station 71, copy and paste the Mean value from the summary spreadsheet to the Mean, Minutes field in the attribute table for Station 71 in ArcMap. Repeat for the Run Count; Maximum, Minutes; Minimum, Minutes; and Standard Deviation fields. Gather the same statistical data for Stations 74 and 76. To generate statistics the eight-minute run data for each station, simply click the Options tab on the table for each station and chose Switch Selection. Copy and paste the eightminute run statistics into the spreadsheet. Now that the time Runs for three District 37 fire stations have been summarized, inspect RunSummary1.xls. Notice that the mean values for all four- and eight-minute runs are very close to the target times. In each station's travel area, the average actual time is very close to the targets. Each station's dataset includes records that represent times greater than and less than the targets. Smaller standard deviation values suggest tighter clustering around the mean. In general, District 37 managers are very satisfied with the performance of this travel network. These trial runs have shown that slightly liberal turn rules and posted speed are valid measures of emergency response capability. Summary This exercise used the Network Analyst OD Cost Matrix to compare actual measured travel times to modeled values. Overall, this analysis demonstrated a very good correlation between modeled and actual times. District 37 personnel continue assessing travel times and testing additional details that might improve the response model. As a corollary to this exercise, District 37 has also recognized the potential to use an OD cost matrix to test and validate actual travel times to specific incidents. Acknowledgments Special thanks go to the administration and staff of Fire District 37, South King County, Washington, and the many FD 37 and Kent City staff who provided data for this exercise. This allowed the mapping and evaluation of actual run time data captured by district firefighters and modeled real travel time data on city streets. To quickly tabulate and display statistics for each station's four-minute runs, locate and rightclick on each table's Total_Minutes field and choose Statistics. To quickly post this summary data, open the Excel spreadsheet template, RunSummary1.xls, provided in the sample dataset and copy values from the four-minute and eight-minute statistics generated for each station to the appropriate cells in the spreadsheet. 54 ArcUser Spring 2009 www.esri.com