The standard has existed for years. The technology has also been around for years. What was missing was a way to make the two halves of a broken system move together. A new FCC order just supplied it.
Nobody wants to call 911. But when they do, they want it to work.
911 is one of the best-known brands out there. Everybody knows the number, but almost nobody thinks about what happens after they dial it, and they shouldn’t have to. From the caller’s point of view, emergency calling is the most critical infrastructure there is. It should just work.
Now here’s what most people never find out: in much of the country, the call you place on a modern smartphone still lands on a 911 network running old analog telephone technology. The idea behind Next Generation 911 (NG911) is to finally rip out that legacy analog infrastructure and put up-to-date, IP-based technology in its place. We’ve had a standard for that for more than 15 years. So why has it taken this long?
The holdup wasn’t the technology itself, it’s about who owns which piece of the system.
A four-link chain, and only half of it is regulated
Think of a 911 call as a chain with four links. It starts with a device, a phone of some sort, that dials 911. From there the call travels to a communications network (AKA a “carrier”), then to a 911 network, and finally to the PSAP, the Public Safety Answering Point where a call taker (911 operator) picks up. Once you dial those three digits, it stops being a normal telephone call. It gets priority routing across the networks. It carries location and other data along with it. That’s how we got the term “Enhanced 911 or E911.”
The FCC regulates half of that chain. It sets the rules for the carriers and the originating service providers (OSPs) that put calls onto the networks. What it cannot regulate is the rest of the chain including the destination: 911 networks and the PSAPs. Those get governed at the local level, and where geography and political boundaries dictate a wide range of 911 authorities. In Ohio and Georgia it’s handled at the county level. In Colorado and North Carolina it’s the state. Georgia went years without any state-level organization at all and is only now catching up.
This is why NG911 stalled. When you can only push from one side of a four-link chain, you get progress in fits and starts. Some states organized early and pulled ahead. Massachusetts and Connecticut are further along than most because they were better organized and started swapping their legacy phone networks for new IP networks sooner. Others are still lining things up. Getting all the stars to align, when only half the ecosystem answers to the same authority, is hard.
What the FCC’s new order actually changed
The new FCC order doesn’t try to regulate the PSAPs. Instead it creates a trigger.
When a local government gets its act together, builds out its NG911 network, and clears a defined checklist, it can formally request service from the originating service providers. Now, instead of delivering calls to the legacy 911 network, the OSPs deliver them to a new IP network, with the required parameters around interconnection and redundancy. And that’s where the teeth are: once a valid request comes in, the OSPs have six months to comply.
NG911 has been talked about for over 15 years, and for most of that stretch the local side and the carrier side had no forcing mechanism between them. Now, the newer NG911 networks have leverage. They can turn to the carriers and say: here you go, deliver the calls here. It’s a good thing this is finally moving, because it lines up incentives that never lined up before.
Two phases, and an honest look at where we are
The rollout comes in two phases.
Phase 1 is IP interconnection. Carriers route voice calls into the new NG911 network, but location still works the old way. This step goes pretty smoothly for a provider that already runs a VoIP network. You tell us where to point, we set up the SIP trunks, and we go.
Phase 2 is where the bigger change lives. Location data starts traveling with the call over those IP connections, and the legacy automatic location information (ALI) databases go away. That one’s harder, and the reason tells you something: every NG911 network is landing on a slightly different specification for how location gets delivered. Give people a choice and they’ll take it, so we’re adapting to a range of implementations instead of one clean standard.
As things stand, Phase 1 has momentum, with over 800 PSAPs having Phase 1 connectivity and active requests for over 1800 more PSAPs being satisfied. Phase 2 is barely off the ground. A handful of requests have come in and then been pulled back, because the local infrastructure wasn’t ready yet. That’s not a knock on anyone. It’s what a big, spread-out transition looks like in its early innings.
The cost of getting the governance wrong
If you want proof that coordination is difficult, look at California.
California runs 911 at the state level and has more than 400 PSAPs to coordinate. The state spent hundreds of millions of dollars trying to build its NG911 system, ran years behind schedule, and toward the end of last year finally admitted it wasn’t working and chose to start over. The director who had been one of the key architects of the effort departed. With a World Cup deadline bearing down, the state had to spend heavily on a stopgap just to get through the near term, and it still has to figure out the lasting fix afterward.
The lesson isn’t that NG911 is impossible. Look at the states doing it well. The common thread is that they settled the governance question first and treated the technology as the second problem, not the first.
What this means if you route calls for a living
The compliance window is here, and the requests are only going to pick up speed. For any carrier or enterprise sending 911 traffic, the question is no longer whether to make this transition. It’s whether you’re built to absorb it or you’ll be scrambling every time a new PSAP comes online.
That’s the position we’ve taken at Bandwidth. Our customers connect to us once (redundantly), to our network edge gateways. Instead of interconnecting with hundreds of separate 911 networks, legacy or NG911, they connect to The Bandwidth network and we aggregate and deliver those calls across legacy and NG911 networks—and as new IP-based networks come online, we handle the back-end work of connecting to them. As we rework the back end to route to shiny new IP networks instead of legacy selective routers, the way our customers provision data and send us calls doesn’t change. We take on the complexity so they don’t have to, which means their compliance work becomes far more manageable as the ground keeps shifting underneath..
NG911 has been fifteen years in the “almost.” This time the incentives finally point the same direction, and the path is clearing for coordinated implementation.
P.S. I spoke about this a while ago on a webinar with my colleague Greg Rogers, an expert in all things telecommunications public policy. You can watch it here.
