Never enter blind.

See with WARDEN. Through smoke, through walls, without GPS.

Nobody can see past the wall.

A fire captain tracks an interior crew by radio roll call. An entry team goes through a door knowing only what it saw from outside. Both are guessing, and the guess is what kills people.

GPS dies indoors

Accountability for who is on which floor falls back to voice, memory, and a status board.

Optical dies in smoke

Cameras, LiDAR, and the eye all quit in a working fire. Thermal sees further but stops at the first solid surface.

Prediction is not detection

Tracking a target from outside the building predicts where it went. If someone is already inside and still, that prediction is worth nothing.

Now somebody can.

WARDEN is one package: radar, UWB and imaging fused into a single position estimate. Radar and UWB each run on the bench today, and fusing them is the current work, so everything below is the finished system rather than what ships now.

Zero-setup positioning

Nothing surveyed, nothing wired. Nodes are placed on arrival and work out their own geometry, inside the window a crew actually has.

UWB self-geometry · no surveyed infrastructure

Tagged and untagged tracking

Every tagged crew member tracked through walls, plus untagged people carrying nothing at all, including anyone lying still.

Radar and UWB fusion

Multi-sensor confirmation

A detection has to be seen by more than one physics before it reaches the commander.

Rule-based arbitration

Live interior picture

Positions rendered against structure geometry rather than a blank grid, so a commander reads rooms instead of dots.

LiDAR-SLAM mapping · open streaming API

Where the physics stops us

Penetration trades directly against resolution. The high-band sensing that resolves a person cleanly through drywall does not meaningfully pass through brick or reinforced concrete. Physics sets that limit, not our software. The finished package pairs both bands rather than betting on one.

Everyone else picked a half.

The tools in this space are mature and they work. Each one solves half the problem, and the halves do not compose: two vendors, two devices, two coordinate frames.

Interior sensing systems compared by technology, whether each tracks tagged crew, whether each detects untagged occupants, and how each is deployed. Only WARDEN does both, and it is the only one that does not require a person already at the structure.
SystemTechnologyTagged crewUntaggedDeployment
WARDENIn development Radar, UWB and imaging fused into one estimate Yes Yes Emplaced or airborne, before arrival
Camero-TechXaver 1000 Pulse UWB radar, AI tracking. 42m, through reinforced concrete No Yes 91cm panel, carried to the wall
LumineyeHandheld radar Pulsed radar, handheld No Yes Handheld, held against the wall
TRX SystemsNEON Inertial, magnetic, barometric and UWB fusion. Patented, sub-metre, floor-level Yes No Worn by each responder
DFR platformsSkydio, BRINC, Flock Optical and thermal from the air, CAD-dispatched No No Airborne, exterior only

What a competitor cannot buy

Every module in this stack is something a competitor can buy, and several of them build excellent products from it. What nobody sells is one frame: tagged crew and untagged occupants in a single picture, from a system already in place before anyone reaches the door. The arbitration behind that is the hard problem, and it is the only part that compounds, because deployments sharpen it program by program.

From bench to burning building.

Phase 1In development

Proof of concept

One position estimate from radar and UWB together, proven end to end on a ground prototype.

Phase 2Planned

Full sensor package, field-ready

The bench system becomes a deployable one. The remaining modalities join the RF core, and the whole package fits a handheld unit with no laptop in the field.

Phase 3Planned

Body-worn sensing

A radar carried on the operator, pointed wherever they look, fused as another moving node. The sensor's own motion becomes a ghost-suppression advantage, which de-risks the airborne work directly. Deployments produce the labelled data a trained classifier needs, held within its own program.

Phase 4Planned

Airborne, GPS-denied

Drones positioned around a floor, sensors looking inward, ranging against each other so drift never accumulates. Triangulation from opposing corners, streamed to a ground station.

Two buyers, one box.

One problem, two buyers, two programs that never touch. They share the engineering and nothing else: separate deployments, separate data, separate delivery. The commercial side hardens the hardware without the government side ever depending on it.

Fire & EMS

Agencies already run Drone as First Responder (DFR) programs. Those aircraft arrive first and see only the outside. WARDEN supplies the inside.

  • Live position of every interior crew member, no GPS
  • Unlocated occupant detection through walls
  • Throw-and-walk deployment
Near-term revenue · DFR and public safety grant pathways

Defense & Intelligence

Room clearing and subterranean operations need the same measurement: who is behind this wall, right now.

  • Pre-entry interior picture for close-quarters operations
  • GPS-denied and subterranean positioning
  • ITAR-walled program, separate data and delivery
SBIR and AFWERX pathways

Emergency payload delivery

The aircraft that carries our sensors can carry equipment. An AED or blood products reach a patient ahead of the ambulance, flown by the agency already operating the package. The flying is proven; the gate is BVLOS authorization.

Payload delivery only. We are not pursuing patient transport.

How we sell into both

We are not building the command platform. WARDEN is software-agnostic hardware with an open streaming API, publishing into the systems our customers already run: TAK on the public safety side, Lattice-class on defense. We supply the measurement and the incumbent owns the screen.

Meet the team.

Guruansh Kohli

Guruansh Kohli

Co-Founder & CEO

CS and Business, University of Rochester. Authored the dual-use thesis behind WARDEN. Leads capital strategy and government pathways.

guruansh.kohli@dronegurus.info
Nate Bogdan

Nate Bogdan

Co-Founder & COO

University of Rochester. FAA Part 107 certified remote pilot. Leads flight operations, field execution, and regulatory readiness.

nate.bogdan@dronegurus.info
Spencer Johnson

Spencer Johnson

Co-Founder & CFO

BS Finance, University of Rochester. Former search-fund manager, 500-plus targets sourced. Leads financial modeling.

spencer@dronegurus.info
Robert Wong

Robert Wong

Co-Founder & CTO

CS, Stony Brook University. Builds the sensing stack: radar signal processing, UWB positioning, the estimator, and the operator interface.

robert.wong@dronegurus.info

Talk to us.

Investor introductions, partnerships, and agency pilots all start the same way. Tell us which one you are and we will follow up.

HeadquartersNortheast U.S. · Delaware C-Corporation

Inquiries are delivered straight to the DroneGurus inbox. Nothing is sent until you submit.