When the Network Bends: The Resilience Audit Every 911 Center Owes Itself This Season

Key Takeaways:

• We can learn from failures. A single cut fiber line in the spring of 2023 put four 911 centers on an unmonitored backup for two days. From this we find three lessons every director can act on: path diversity must be tested, backups must be monitored, and resilience must be the standard at all times. 
• Interoperability lives in the i3 standard and lags in practice. Vendor silos, version mismatches, GIS boundary gaps, and missing mutual-aid governance are where otherwise-resilient networks bend past their limits.
• Eight checks for hurricane season. The resilience checklist INDIGITAL’s Sandi Stroud presented at NENA 2026 is a network-level audit a 911 center can start ASAP, before this hurricane season highlights the gaps in your network for you.

 

It’s mission critical to maintain a flexible network as we head into hurricane season. While Accuweather may have downgraded its forecast from 11 to 16 named storms to 8 to 14, it still expects as many as seven hurricanes (four of which could be major) and three storms could directly impact the U.S. At particular risk are the northern and eastern Gulf Coast, along with the southeastern coastline and parts of the Caribbean. If a super El Niño doesn’t materialize, expect that forecast to increase. If your ECC isn’t prepared for the coming storms, or if your network is too rigid, there could be gaps in your operation (and you won’t find them until your storm touches down). 

Think of legacy routing methods as oak trees. They’re tough and sturdy to look at, but when too much pressure is applied to one point, it snaps as if it were a twig. A good 911 network is supposed to bend, not break. We see this in engineering all the time, where flexibility is built into buildings so they sway by design, or in aircraft wings so they can flex through turbulence. For 911, readiness means having the capacity to bend. It’s why next-generation 911 (NG911) was drawn up, to distribute the load across diverse paths into an ESInet in place of a single circuit. At the same time, policy-based routing dynamically transfers calls between 911 centers when one is overwhelmed, while a geodiverse ECRF and LVF mean that location and routing never share a common point of failure. Do you know how flexible your network is?

Last month, we wrote about three real events that tested network flexibility, including the Nashville bombing, Hurricane Helene, and Hurricane Milton. This piece is about the other side of the same coin. As we enter the early weeks of hurricane season, what is your network assuming that nobody has tested?

Simple Failures Cause Major Problems

When a rare early-season tornado hit the upper Midwest in the spring of 2023, a fiber line failed and caused wide-scale damage. When the fiber went down, calls that failed were carried over to a secondary path already contracted by the ECC. It hobbled four 911 centers, forcing them to spend the next two days running on a backup. But the provider of that network did not have a network operations center (NOC) or 24/7 support, which means those four ECCs were running unmonitored until the fiber was fixed. This was entirely preventable, yet it exposed an operational risk to be solved. It could be much worse, which is why the three lessons learned on the other side of it are so critical:

Lack of Diversity: An unseen, untested singular point of failure a state away shut down four 911 centers. If your path diversity hasn’t undergone end-to-end testing, then you’re just waiting on a storm to test it for you.

A backup you can’t support isn’t a backup. The secondary path did its job and let calls flow. But for two days, nobody was watching it. A failover with no NOC behind it isn’t risk-free. Rather, it’s passed the risk on elsewhere.

Procurement became the operational risk. The resilience decisions had been made at contract time, years earlier, and never revisited. Don’t wait for your storm to uncover what procurement overlooked.

This is not just a rural Midwest story either. We’re now seeing this pattern playing out on a national scale. For most agencies, the cause of failure isn’t a storm but simple human error. A few years back, the same root cause from the Midwest outage resulted in a multi-state network failure, disrupting 911 for hours as workers installing a light pole in Missouri clipped a fiber line and knocked out 911 service for ECCs in Nebraska, Nevada, and South Dakota. Meanwhile, people calling 911 in Del Rio, Texas, were put at risk when an outage affected call deliverability. The failure in Del Rio wasn’t due to the city’s 911 system, but a downed network from a cellular carrier. The FCC then investigated the multi-state outages, issuing a statement: “When you call 911 in an emergency, it is vital that the call goes through.”

It’s worth noticing what these events have in common with each other: human error and a single point of failure. For this reason, the weeks before peak hurricane season are the right time to test the gaps listed below.

The Interoperability Imperative

The reasons your network fails are numerous, but they don’t make themselves obvious on any given day. Rather, they cause failure at exactly the moments you need cross-agency transfer and shared routing to just work. Critically, the i3 standard anticipates these failure gaps. The problem is that standard practice is not as forward-looking as it needs to be. So, at NENA 2026 in the Innovation Theater, INDIGITAL’s Vice President of Public Safety, Sandi Stroud, named four gaps to pay attention to:

Gap #1: Vendor silos. Neighboring agencies on different ESInet vendors that cannot natively transfer calls or share routing. The mutual aid the architecture inherently promises vanishes when there’s inconsistent implementations between systems.

Gap #2: i3 version mismatch. Inconsistent implementations are like trying to speak a sentence in English using a different dialect for each word. It’s the same standard (English) but the differences are stark enough to be incomprehensible.

Gap #3: GIS boundary gaps. Mismatched boundaries between agencies can send calls to the wrong center, and the consequences of this range from a simple misroute to a compounding problem during a large-scale emergency.

Gap #4: The governance gap. The widest gap of the four. This is where no mutual-aid routing was established in advance, causing directors to improvise on the fly rather than have a plan in place when they need it.

The bottom line here is that interoperability lives in the standard. Getting it to live in practice is agreement work, boundary work, and paperwork. In other words, it’s fair-weather work where the benefits are made crystal clear to those who finish it come hurricane season.

Your Disaster Resiliency Checklist

Stroud closed out her NENA talk with a disaster resiliency checklist you can start immediately. If last month’s blog was about the written COOP every center needs, this checklist is about the network assumptions a COOP relies on. Afterall, a cutover plan is only as good as the paths it cuts over to. Here are eight items to check off your list sooner rather than later:

1. Map every single point of failure: carrier, ingress, power, backhaul, GIS. If the incidents in this blog teach you one thing, it’s that your ECC’s point of failure may be a state away.
2. Confirm true path diversity: not two circuits on the same provider.
3. Run a live failover drill this season: Don’t just run a tabletop. Only a live drill can test your emergency preparedness
4. Verify your backup path has a NOC and 24/7 support: If nobody’s watching the failover, it has already half-failed.
5. Sign cross-jurisdiction transfer and mutual-aid MOUs before you need them: mid-event is the most expensive time to negotiate.
6. Check GIS boundary accuracy and ECRF/LVF currency: routing is only as good as the boundaries it routes by.
7. Establish a comms fallback for when CAD or admin systems fail but the call path holds: the call arriving is half the job; coordinating the response is the other half.
8. Set a recurring resilience review, owned by a named person: discipline without an owner is not discipline.

 

If you run this audit now, in July, you might spend a few days running your team more than usual. If you run it during an emergency event, you’ll be doing the same but with live callers in crisis on the line. And when peak season is here, every one of them gets harder to test, and some become impossible once the first hurricane makes landfall. 

911 Resiliency in Action

Readiness is developed on calm days. A 911 center that treats storm season as a discipline rather than an event is a center whose worst day looks, from the caller’s side, remarkably like any other. That is, the call connects, someone answers, and help is dispatched to the scene.

That is the standard INDIGITAL builds toward, and the reason resilience sits at the center of our work rather than the margin of it. When local systems fail, MEVO services ( INDIGITAL’s backup call-handling system) keep 911 answering across fiber, LTE, and Starlink, an approach Frost & Sullivan recognized in 2026 with its Product Leadership Award for Disaster Recovery Solutions. In fact, INDIGITAL is the only company named in the category this year. This is the entire point of what we do: connecting calls on your worst day.

Frequently Asked Questions (FAQ)

What is true path diversity in a 911 network?

True path diversity means emergency call traffic has two or more fully independent routes into the network – different carriers, different physical conduits, different geography – traced and verified end to end. Two circuits that share a provider, a conduit, or an upstream interconnection point are a single path with redundant billing. The 2023 upper Midwest incident, in which one cut fiber line put four 911 centers on an unmonitored backup for two days, traced to a shared point of failure a state away that no local diagram showed.

Why do NG911 systems from different vendors struggle to interoperate?

The NENA i3 standard defines how NG911 systems should exchange calls, location, and routing data, but implementations vary by vendor and by version. In practice, neighboring agencies on different ESInet vendors often can’t natively transfer calls or share routing policy, and mismatched GIS boundaries between agencies can misroute calls. Closing the gap is largely governance work: cross-jurisdiction transfer agreements, mutual-aid MOUs, and boundary reconciliation completed before an event requires them.

What should a 911 center do first to prepare for hurricane season?

Start with a single-point-of-failure map covering carrier, ingress, power, backhaul, and GIS dependencies – most of the remaining work falls out of what that map reveals. From there: verify path diversity end to end, run a live failover drill rather than a tabletop, confirm the backup path has 24/7 NOC support, and put mutual-aid MOUs in writing. Each item is achievable in fair weather and materially harder once a storm is on the map.