Editing Computer-aided dispatch (section)

==Service levels and geographic information==

[[Geographic information system|Computerized mapping]], [[automatic vehicle location]], [[automatic number identification]] and [[Caller ID|caller-identification technology]] are often used to enhance the service by pinpointing the locations of both the client and the most suitable vehicle for serving the client.

Some CAD systems allow several sources of information to be combined. For example, adding automatic vehicle location (AVL) and geographic information (GIS) could improve service by getting units to a service call location faster. Ideally, CAD is connected to monitor vehicle locations provided by an AVL system. This information is used to suggest the closest vehicle to an event. How is the closest unit determined?

===Basic zone system===

The simplest system is a beat or zone map system. For example, in a community with four fire stations, a grid is overlaid on a community map. Each zone of the grid is identified with a progression of police beats, ambulance zones, transit zones, or fire stations.<ref>This would work for any system including taxis or parcel pick up.</ref> One grid might be labeled: AB241. This means fire station 2, then 4, then 1, then 3 would respond to a fire call occurring inside this zone. The predefined order is created by persons with expertise in the service being provided, local geography, traffic, and patterns in calls for service. 

Since only basic GIS information is included, if AVL was available, it would simply display service vehicle locations on a map. The closest unit would be interpreted by the dispatcher looking at vehicle locations projected on the map.

Where detailed geographic data are not available, units may be assigned based on the center of a district. To make the computing problem easier, the CAD system may use ''[[Centroid|centroids]]'' to evaluate service vehicle locations. Centroids are estimated center points within a zone. The system calculates a distance from a fire station or AVL location to a centroid point. The closest fire station, according to CAD system rules, would be assigned. Systems may use centroids that are not exactly centered in order to skew or weight system decisions. Staff based at a fire station that is physically closer by drawing a straight line on the map may be slower to reach a zone. This can occur because responding units must drive around freeways, lakes, or terrain obstructions in order to reach a zone. A centroid may be moved because 200-car freight trains often block a railroad crossing used to access a particular zone.

This is the cheapest system to develop because it requires the least detailed geographic information and the simplest calculations. Another problem occurs where several services use the same system. Police and transit, for example, may have different ideas about what boundaries define the ideal zone or how centroids should be weighted.

=== CAD using geocoding===

[[Geocoding]] is a [[translation]] system allowing addresses to be converted to X- and Y-coordinates. Someone placing a call for service has an address attached to a wired phone number or tells the dispatcher their address. For example, suppose the caller's address is ''123 Main Street''. 

The GIS or CAD system includes a look-up table. The table may identify odd-numbered addresses in the community as being on the north and east sides of streets. Addresses from 113 to 157 Main Street are identified as being along Main Street's center line between Broadway and Washington. 123 is estimated to be on the north side of Main Street somewhere closer to 113 than 157. This estimate produces a [[latitude]] and [[longitude]], or a set of [[Universal Transverse Mercator]] coordinates. The coordinates are close enough to identify the closest service vehicle. This system may automatically append the name of the nearest cross-street or intersecting street.

Again, the system uses a straight-line distance to determine which service vehicle is closest to a call for service. If an AVL system is used, the CAD system will look through a list of most recent reported vehicle positions. Next, the positions are compared to the service vehicle status. The CAD system may identify several of the closest units that have a status of ''available''. The dispatcher makes an ideal choice from the CAD system shortlist.

This type of system is significantly more expensive than a zone system. The basic system may start with maps from the US Census Bureau or a county assessor's office. The quality of these maps may be good but will not be ideal for dispatching. There would normally be one or more persons on staff who would deal with data changes from new development, new streets, or data quality problems. The person would compile addresses and generate street centerlines in mapping software. Geocoding varies in accuracy depending on data sources and vendors. It normally takes years of work and planning before a system is implemented. Modern geocoded systems will often display service vehicle locations, the location of service calls, and the locations of callers on a map. This helps to disambiguate calls for service and reduces the likelihood of dispatching two reports of a single call for service as two separate calls.

Another problem comes from technologies using differing datums or coordinate systems. For example, suppose your AVL system uses degrees-decimal degrees format. The AVL display for a vehicle at the Heart Butte Post Office in Montana shows a latitude and longitude of 48.28333 N, -112.83583 W. The CAD system uses degrees-minutes-seconds format data and shows the same location as 481700N, 1125009W. How do you translate? This is sometimes a problem with neighboring CAD systems. Ideally, you should be able to send and receive calls to and from CAD systems in neighboring areas. What if the state or provincial government has standardized on a different coordinate system?

===Full GIS/AVL integration===

The most expensive and technically challenging systems fully utilize the capabilities of geographic information systems (GIS) and automatic vehicle location (AVL). In these systems, the street centerlines are described as ''routable''. In addition to geocoding and accurate street centerlines, intersections have attributes or scores. Can a service vehicle turn left from eastbound Carnegie Street onto northbound Hooligan Boulevard? A scoring system is used to assess the difficulty of making the turn. At one end of the scoring system there might be an interchange where service vehicles had unrestricted access in making the turn. Perhaps both streets are one-way, making it relatively easy to turn from one onto another. In the middle scores, a left turn might be blocked occasionally by heavy traffic, a [[draw bridge]], or [[street cars]]. At the most difficult score, the two streets may cross but the lack of any interchange does not allow service vehicles to get from one to the other.

To calculate the closest service vehicles, the CAD system does a network analysis of the road system based on these routable street centerlines. It assesses the path from the service call to the AVL location of available vehicles. The system recommends the service vehicles with the shortest path. 

Routable street centerlines take into account differences between northbound and southbound lanes on a freeway or turnpike. For example, to reach a point in the southbound lanes of a turnpike, service vehicles may need to drive north to the next exit then return on the southbound side. The analysis of a routable street network takes this into account so long as the event location is accurately reported. Routable systems account for barriers like lakes by calculating the distance of the driven route rather than a straight-line distance. It is assumed the service vehicle driver knows the shortest path or that all drivers make similar numbers of wrong turns.

===Concentration===
CAD systems require support staff with special skills. This can lead to concentration of dispatch facilities, particularly where there is population growth or where automation is required to meet defined service objectives. 

In any system, concentration of facilities increases risks of outages or massive failures. In a system where the call traffic is so high that advanced technology is needed to handle routine levels of day-to-day calls, relatively minor failures can have major effects on service levels. For example, where everyone is used to the convenience of automatic vehicle location (AVL), an AVL outage can suddenly increase staff workloads. Suppose a failure causes a condition where CAD cannot recommend a closest unit. How will the dispatcher efficiently assess which unit to assign?