January 13, 2025
Auto dispatch is one of those things that sounds obvious once you see it in action, but I’m not sure how many CAD vendors even offer it or how many agencies are taking full advantage of it. Spoiler alert: they should. It’s a perfect example of how technology can save time—not just any time, but time at the most critical part of the incident, right at the start.
I’ve written about how time is the enemy in emergency services here: https://www.linkedin.com/pulse/time-keeps-ticking-our-battle-against-jonathan-boyd-aahjc/?trackingId=qw0E6H1sToKEsT9YCUMjDw%3D%3D
Our department has been using auto dispatch since 2014, thanks to CentralSquare Technologies CAD. It works much like it sounds: give the CAD some basic info, and it dispatches units, automatically. The reality is a little more intricate. When a 911 call comes in, the dispatcher validates the location and grabs basic details. Once they select a call type, fire or EMS, the CAD determines the closest units using Automatic Vehicle Location (AVL), dispatches them, and sends alerts to station software. This triggers everything: station speakers, radios, and mobile data computers (MDCs). The whole process takes less than three seconds.
Before auto dispatch, this process was manual and took an average of 1 minute and 55 seconds. Now we’re at 22 seconds. That’s over a minute and a half saved—per call. When you’re “hunting for seconds,” that is a lot of seconds. But auto dispatch isn’t a magic wand; there are a few things you have to figure out to make it work.
Response Practice Considerations
The key to auto dispatch is getting the call type selected quickly, which means simplifying your options. Agencies that use priority dispatch or tiered responses might struggle here. If you’re spending extra time deciding whether an EMS call is ALS or BLS, you’re eating into the time savings. A simpler approach, basically deciding if the call is fire or EMS, gets units rolling faster, but this oversimplification may not work for everyone.
Another issue: errors happen. Dispatchers occasionally select the wrong call type. If “suspicious person” is right next to “seizure” in your system, mistakes will be made. Most of the time, it’s harmless (units get canceled, no big deal), but it’s something to think about if you share dispatch services with police. The real question is whether saving time upfront is worth the occasional error and whether those seconds saved can resolve incidents faster, returning units to service sooner.
GIS Data: The Backbone of Auto Dispatch
GIS data is the cornerstone of auto dispatch. AVL uses GIS data to calculate the closest unit based on route time. If your GIS data isn’t perfect, the whole system falls apart. Common issues include disconnected street segments, incorrect elevation data, duplicate street names, bad segment speeds, and wrong street directions. For example, a highway flyover with bad elevation data might make the CAD think a unit can leap off the overpass onto the road below. Duplicate street names can confuse address validation, while incorrect street directions make the CAD think a unit can drive the wrong way on a divided road.
Street segment speeds are another problem. Many agencies enter speed limits, but real-world travel speeds for emergency vehicles are often faster (or slower). The best approach is to estimate realistic speeds based on your units.
Maintaining accurate GIS data isn’t just about fixing issues at installation. Street networks, intersections, and conditions change over time. Construction projects, new developments, and road closures can make your GIS data obsolete fast. Without a quick and reliable process to update data, errors will creep in and erode trust in the system.
GIS Tools That Can Help
Esri provides several tools to address these GIS challenges. These include:
- ArcGIS Data Reviewer: Automates data validation to find errors like disconnected streets, duplicate names, or incorrect elevation data. It provides batch processing and configurable checks, making it easier to spot problems before they cause dispatch errors. https://www.esri.com/en-us/arcgis/products/arcgis-data-reviewer/overview
- ArcGIS Network Analyst: Ensures accurate AVL routing by analyzing service areas, closest facilities, and route optimization. It allows adjustments for travel speeds and custom road conditions to improve dispatch recommendations. https://www.esri.com/en-us/arcgis/products/arcgis-network-analyst/overview
- Geometric Network Tools: Help visualize and troubleshoot connectivity issues to ensure roads and intersections are mapped correctly. These tools are essential for identifying disconnected street segments or errors in routing logic.
Even with these tools, there will always be gremlins. We tested our system with over 3,000 calls before launching and still missed some. That’s why it’s important to think of GIS maintenance as an ongoing task, not a one-time project.
Simulation Testing: Building Trust in Auto Dispatch
Trust doesn’t happen overnight, especially when you’re introducing a system as complex as auto dispatch. Like most relationships, trust takes time, repetition, and consistency. In our “zero fail” industry, that trust has to be earned before the system goes live.
This is where simulation comes in. Before we implemented auto dispatch, we ran over 3,000 test calls in just five days. These weren’t random scenarios, they were designed to simulate real-life conditions, including common errors like incorrect call types or tricky GIS situations. Running these tests gave us the “sets and reps” needed to build confidence in the system.
Here is an article I wrote about how the military uses simulation to build trust in technology: https://www.linkedin.com/pulse/how-fire-service-can-learn-from-military-simulation-training-boyd-hcrrf/?trackingId=bzokPD%2BqTAuE%2BPR3FLXWnQ%3D%3D
Simulation testing isn’t just about finding errors (though it’s great for that). It’s also about preparing your team. Dispatchers, firefighters, and officers need to see how the system works in action. When they understand the “why” behind auto dispatch and see it in action, they’re more likely to trust it when it’s live.
Another way we built trust was through transparency. We set up a 24/7 helpline so firefighters could call in and ask why a particular unit was dispatched. This stopped rumors before they could spread. Without clear communication, rumors become truth, and trust in the system crumbles. The helpline also helped us identify and fix legitimate issues, like GIS errors that slipped through during testing.
The Future of Auto Dispatch
Auto dispatch has come a long way, but there’s still room to grow. Real-time data integration—like live traffic conditions, weather, and road closures—could make dispatch recommendations even more accurate. Artificial intelligence is another frontier. Some systems are already testing AI tools that “listen” to 911 callers to help determine call types faster.
After more than 10 years, our system is still saving over a minute on every emergency call. But getting here wasn’t easy. Auto dispatch isn’t a “plug and play” solution. It takes preparation, adjustments to response practices, GIS maintenance, and a commitment to ongoing testing and trust-building.
I’m still not sure why every CAD vendor doesn’t offer auto dispatch or why more agencies aren’t using it. Maybe there’s not much market demand. But if you’re serious about finding time in your process, it’s worth a look. If your agency has implemented auto dispatch, I’d love to hear how it’s working for you.