Core Aerobots builds self-flying drone swarms that sweep entire disaster zones in minutes — and use onboard AI to find the people who can't wait. One operator. A whole fleet. Day or night.
On July 4, 2025, the Guadalupe River rose more than 20 feet in under an hour. By the time rescuers reached the worst-hit stretches of the Texas Hill Country, the water had already taken more than a hundred lives — many of them children.
The rescuers weren't slow. They were heroic. There simply weren't enough eyes covering enough ground, fast enough. Ground teams search in straight lines. Helicopters are few, expensive, and mostly grounded at night — and night is exactly when the water rose.
The first hours are the ones that decide who survives. Too many of them were spent looking, not finding.
We started Core Aerobots to close that gap — not with one more aircraft, but with a swarm. Dozens of low-cost drones that launch together, divide the search area among themselves, and scan it with thermal and visual AI, by day or by night. What takes a ground team hours, the swarm covers in minutes. Every drone thinks as part of the whole, so one operator can search like a hundred.
Draw an area and a safety boundary on the map. That's the whole input — no flight plan, no manual piloting.
Live todayDrones arm, take off, elect a leader, and split the zone between themselves — coordinating over a resilient mesh with no single point of failure.
Live todayEach drone scans with thermal and visual detection on an onboard AI accelerator, cross-checking sightings across the fleet to kill false alarms.
Funded nextLive video and positions stream to the operator in real time; five safety layers keep every aircraft inside the lines and bring it home on low battery.
Live todayDistributed coordination lets dozens of drones fly, divide the work, and decide as one unified system — leader election, task allocation, and formation control built in.
Core PlatformOnboard neural inference turns raw camera feeds into found people — running on a Hailo-8L AI accelerator at the edge, with no network connection required.
PerceptionThermal imaging finds body heat through darkness, smoke, and tree cover — so the search never stops when the sun goes down or conventional aircraft are grounded.
SensorsClassifies what it sees — person, vehicle, hazard — and the swarm votes across overlapping fields of view to confirm a detection before it ever reaches the operator.
IntelligenceHigh-level command authority at fleet scale: arm, search here, return all, land — identical for one drone or a hundred. Operators keep manual override at every moment.
OperationsPrecision waypoint navigation inside keep-in / keep-out boundaries the aircraft physically cannot cross. The fleet follows the plan and stays within the lines — always.
NavigationEncrypted, low-latency video and telemetry from every drone to a single command view — see what the entire swarm sees, live, as it happens.
Data LinkVehicle-to-vehicle mesh keeps the swarm coordinated even when individual links drop. No central tower to knock out — the network heals around losses.
NetworkingReal-time proximity detection between drones and obstacles, with an ORCA avoidance engine — so the fleet can fly dense without flying into itself.
SafetyFlying over an area isn't the same as searching it. A camera no one is watching finds no one. Core Aerobots drones don't just fly the grid — they understand what's below them.
Onboard AI runs thermal and visual detection at the edge, picking out the heat signature of a person against cold floodwater or dark forest, then confirming it across the swarm's overlapping views before flagging it to the operator.
A last-resort gate that validates every setpoint — altitude, speed, boundary — before it can ever reach the flight controller.
20 Hz containment monitoring with keep-in / keep-out zones the aircraft cannot cross.
Priority-based failsafes — battery, link loss, motor fault — mapped to safe actions like return-to-home or immediate land.
A nine-mode state machine that refuses unsafe transitions — no arming, no takeoff, unless conditions are met.
Live proximity-threat detection between drones and obstacles, with real-time deflection.
A multi-drone swarm flies full autonomous missions with a five-layer safety stack — validated in high-fidelity simulation against real flight-controller software.
Target identification on the onboard AI accelerator, production hardening, and flight testing on real aircraft. This is what the next dollars build.
Deployment alongside first-responder agencies for flood and wilderness SAR — then wildfire response, security, and infrastructure.
We're raising to take a proven autonomous swarm from simulation to the field — and put it to work where minutes decide who survives.
Sweep submerged neighborhoods and cut-off areas in minutes, with thermal detection that works at night — when conventional aircraft can't fly. Live feeds give incident command eyes everywhere at once.
Cover vast, roadless terrain a ground team would take days to walk. The swarm divides the grid and prioritizes high-probability zones automatically, scanning by thermal and visual at once.
Map fire lines and spot people and crews through smoke with thermal imaging, feeding a live common operating picture to commanders — without risking a crewed aircraft in the plume.
The same autonomous swarm and safety stack extend to persistent overwatch, perimeter security, and infrastructure inspection — at a fraction of the cost and risk of crewed assets.
Core Aerobots partners with first-responder agencies, government, and investors to bring autonomous search-and-rescue swarms to the field. Reach out to schedule a live demonstration or to discuss partnership and investment.