RTT in 911: Why Real-Time Text is Becoming Critical Infrastructure

When someone with hearing loss dials 911, they face a brutal choice: struggle through a voice call or waste precious seconds typing an email to a non-emergency line. RTT—Real-Time Text—is closing that gap, but the rollout is messy, and most engineers building backend infrastructure don't fully grasp its criticality.

Let's be direct: RTT isn't a nice-to-have accessibility feature. It's essential infrastructure that directly impacts whether people survive emergencies.

What RTT Actually Is (And Why It's Not Just Texting)

Real-Time Text transmits text character-by-character as it's typed, not after you hit send. This matters because traditional SMS to 911 works like email—you type, you submit, then dispatch sees it. With RTT, the dispatcher watches text appear in real-time, exactly like you're having a conversation.

This distinction sounds minor but it's operationally massive. Dispatch can ask clarifying questions mid-sentence. A person having a stroke can begin typing "I can't move my left—" and the dispatcher already starts coordinating response. The latency is milliseconds, not the seconds or minutes SMS introduces.

The Technical Layer

RTT runs over IP networks, typically integrated into VoIP infrastructure. Most 911 centers (PSAP—Public Safety Answering Points) are still analog-heavy. Retrofitting them to handle simultaneous voice and character-level text streams requires upgrades that many rural and mid-sized departments simply haven't funded.

The challenge isn't the protocol—it's the deployment. You need call routing that understands RTT, logging systems that capture character sequences, and dispatcher consoles that render text in real-time without dropping characters. One missed keystroke in a medical emergency is unacceptable.

Why the Urgency Is Now

The FCC mandated RTT support in 911 systems by June 2024. That deadline has passed. Some jurisdictions complied cleanly. Many haven't. As of mid-2024, coverage is fragmented—your RTT call might work in San Francisco and fail entirely in rural Pennsylvania.

The Adoption Cliff

Deaf and hard-of-hearing communities spent decades asking for this. Phones have supported RTT since iOS 13 (2019) and modern Android devices. The bottleneck was always 911 infrastructure, not consumer devices. Now that mandates exist, the expectation is that it works—but it doesn't, uniformly.

This creates real liability. If someone calls 911 via RTT and the PSAP isn't equipped to receive it, what happens? The call fails silently or degrades to voice-only, leaving the caller in an impossible situation. That's not a UX problem. That's a life-safety problem.

Integration Complexity

For engineers building backend systems that touch 911—whether you're working on carrier infrastructure, emergency management software, or dispatch tooling—RTT introduces new requirements:

Character-level transmission and receipt logging. Timestamp accuracy for every keystroke, not just the call start. Session state that survives network transitions. Error handling for partial text reception. Accessibility compliance for dispatch interfaces that historically weren't designed for real-time text input.

If you're building dispatch software or working at the carrier level, you're dealing with these constraints whether your client has explicitly asked for RTT or not. The mandate makes it non-negotiable.

Why This Matters for Your Infrastructure

RTT forces a shift in how we think about emergency communications architecture. Voice-first design is no longer acceptable. You need multi-modal systems that can handle simultaneous voice and text, each with their own latency and reliability requirements.

At LavaPi, we've helped clients upgrade 911-adjacent infrastructure to support RTT requirements—both the telecom side and the dispatch software side. The projects aren't flashy, but they're critical. If you're in that space, RTT compliance isn't optional anymore.

The Real Takeaway

RTT is infrastructure, not features. It's a shift from a voice-first emergency system to one that treats text as a first-class communication channel. The mandate exists because deaf people shouldn't have to choose between getting help and using their preferred language. The engineering challenge is making that work at scale, across fragmented, aging 911 infrastructure. That's the problem we need to solve, and it needs to work before the next emergency.