On June 24, 2026, the city of Montevideo did something that public-safety agencies in the United States have talked about for years but have rarely, if ever, fully automated: it cut the human dispatcher out of the loop between a gunshot and the drone that goes to investigate it.
Drone-autonomy firm FlytBase, working with Uruguay's Ministry of the Interior and local partner Timerix, switched on a system that ties Montevideo's existing ShotSpotter acoustic gunshot-detection network directly to a fleet of autonomous drones. The mechanism is deliberately simple to describe and technically hard to do at city scale: when the acoustic sensors register gunfire, the nearest drone launches automatically from its docking station, flies itself to the reported location, and begins streaming live aerial video to a command center — all within minutes, and with no pilot manually taking off, navigating, or babysitting the flight.
FlytBase and its partners are calling it one of the first deployments of its kind in Latin America, and say Uruguay is the first country to integrate automated drone deployment directly with an existing ShotSpotter network at scale. Timerix, the local partner that helped stitch the architecture together, says the setup can "deploy drones in seconds."
What's actually new here
The phrase to watch is "Drone-as-First-Responder," or DFR — a model in which a drone, rather than a patrol car, is frequently the first asset to arrive over an incident and feed situational awareness back to commanders before officers are on scene. DFR is not a new idea, and U.S. departments have run versions of it for several years. What separates Montevideo from most of those programs is where the human sits in the chain.
In a typical DFR operation, a sensor alert or a 911 call still passes through a person — a dispatcher, a remote pilot, a duty officer — who decides to launch and, in many cases, manually flies or supervises the aircraft beyond visual line of sight (BVLOS). Montevideo's deployment closes that detect-to-launch loop without a human dispatcher in the middle. The trigger is the gunshot itself. The ShotSpotter network localizes the acoustic event, and that data drives an automatic dock launch.
That is the technically distinct part. It is the difference between "a sensor tells a human, who launches a drone" and "a sensor launches a drone, and a human watches the video." The first is workflow automation. The second is sensor-triggered autonomy wired into legacy public-safety infrastructure — and it is meaningfully harder to certify, operate, and trust.
How the pieces fit together
Three layers make the system work, and each maps to a different player.
The detection layer is ShotSpotter, the acoustic gunshot-detection network Montevideo has used since 2023. Arrays of microphones triangulate the sound of gunfire and produce a geolocated alert. That infrastructure predates the drone program; the new system treats it as a sensor feed rather than as an end product that simply pings a human operator.
The autonomy and operations layer is FlytBase. Its platform handles the citywide orchestration: coordinating flights across multiple docking stations, managing BVLOS missions, and piping real-time video back to the command center. The hard problem at city scale is not flying one drone — it is deciding which dock launches for a given alert, deconflicting airspace, managing battery and dock state across a network, and keeping a reliable live feed flowing to operators. That coordination layer is what turns a collection of docked drones into a responsive citywide capability.
The integration and local-deployment layer is Timerix, the Uruguayan partner that connected the acoustic network to the autonomy platform on the ground and is responsible for the "deploy in seconds" architecture claim.
On the compliance side, FlytBase says its platform complies with standards including SOC 2 Type II, ISO 27001, and GDPR — the kind of security and data-governance credentials that matter when a government is routing live aerial surveillance of its own citizens through a vendor's software.
Why a country wires sensors straight to drones
The logic behind sensor-triggered drone response is the same logic that has driven the broader expansion of drones across public safety: getting eyes on a scene faster than a ground unit can arrive. Drones have become a serious enough factor in the field that U.S. Department of Justice material from the National Institute of Justice now catalogs the detection technologies — acoustic, radar, and radio-frequency among them — built to identify drones operating where they should not, such as over prisons and jails.
The appeal is straightforward. A drone overhead within a couple of minutes can confirm whether a ShotSpotter alert is a real shooting or a false trigger, count and locate victims, track fleeing suspects or vehicles, and give incoming officers a live picture before they commit. Removing the human launch decision compresses the timeline further: every second spent waiting for a dispatcher to read an alert and push a button is a second the drone is not airborne.
But the operational and legal considerations that come with autonomous public-safety UAS do not disappear because the system is faster. They arguably get sharper.
The questions this raises
Is there really no human in the loop? The detect-to-launch decision is automated — the gunshot triggers the flight. A human is still very much in the loop on the receiving end, watching the live video at the command center and deciding what to do with it. The automation is at the front of the chain, not the back. That distinction matters: the system is autonomous in dispatch, not autonomous in judgment or enforcement.
What about ShotSpotter's accuracy? Acoustic gunshot detection has been controversial in the United States, with disputes over false-positive rates and the value of the alerts it generates. Wiring that sensor directly to an automatic drone launch inherits whatever accuracy — or inaccuracy — the underlying network has. A false acoustic trigger now produces an automatic flight rather than a discarded alert. The flip side is that an autonomous drone overhead is, in principle, a fast and relatively low-cost way to verify or dismiss an alert without committing officers.
Why Uruguay, and why now? Uruguay is a small, comparatively well-governed country, which makes it a plausible proving ground for a city-scale autonomous DFR system without the fragmented jurisdictional and regulatory patchwork a U.S. metro would face. A single national Ministry of the Interior, an already-installed ShotSpotter footprint in Montevideo, and a vendor stack that handles the autonomy and compliance layers together lower the barrier to a clean, end-to-end deployment.
Why It Matters
For most of the DFR era, the gap between "sensor detects something" and "drone is in the air" has been bridged by a person. Montevideo is a concrete proof point that the gap can be closed automatically at city scale, with sensor-triggered, dock-launched, BVLOS autonomy bolted onto public-safety infrastructure that was never designed with drones in mind.
That is significant for two reasons. First, it shifts DFR from a faster dispatch model toward a genuinely automated response model — the kind of architecture other governments will study, and that vendors will market aggressively. Second, it front-loads the hard policy questions. When a gunshot can put a government drone over a city block with no human deciding to send it, the accuracy of the trigger, the oversight of the video feed, the data-governance of the vendor, and the legal framework around automated aerial surveillance all stop being hypothetical. Montevideo is now a live test of whether closing the detect-to-launch loop delivers faster, better public-safety outcomes — or simply automates the parts of the system that civil-liberties advocates were already worried about.
