Navigation· SenzFusion

Independent navigation in contested environments

Jamming and spoofing can take GNSS away. SenzFusion navigates on its own sensors: the IMU predicts motion, radar map matching corrects drift and a thermal scene check confirms the fix, while GNSS input is used to detect a contested environment. The host autopilot receives course corrections and keeps flight control.

IMU motion prediction Radar map match Thermal scene check GNSS · check only
ShoalOS Simulator screen: a 3D alpine valley with a UAV following a waypoint route, beside camera, imaging IR, radar and navigation-sensor panels and an event log.
ShoalOS Simulator · GNSS outage rehearsalScripted demo
Sheet
Navigation · Pipeline
Layers
L0 · L1 · L2
Status
Payload + SDK · demo scripted
Revised
Sheet 01Navigation pipelineSelect a stage

One estimate, two witnesses

The IMU predicts motion throughout the flight. Radar and thermal matches against stored references correct and cross-check the estimate; a position fix is accepted when both agree.

check update along-track + cross-track vs stored terrain map vs reference scene PrimaryIMU · predict motion FPGA SoCContinuous estimate GNSS · check onlyDetects contested area Ground profileRadar altimeter CorrectionRadar map match VerificationThermal scene check Position fixBoth agree InputThermal camera
Primary

Predict motion

The IMU predicts position, velocity and orientation 1,000 times a second, including between fixes. Barometer and airspeed feed the same fusion core.

Why FPGA

Predictable timing. Motion prediction, the radar map match and the thermal scene check run in fixed-timing FPGA logic.

Proposed board architecture
Sheet 02Processing coreFPGA · ARM

Sensor reading to course correction in six steps

SenzFusion runs on an FPGA system-on-chip: fixed-timing logic for the sensor pipeline and ARM cores for the estimate update.

Six processing steps, their inputs and where they run
StepInputWhat happensRuns on
1 · Time-stamp + cleanAll sensorsA hardware clock time-stamps readings; weak readings are droppedFPGA
2 · Predict motionIMUPosition, velocity and orientation predicted 1,000 times a secondFPGA
3 · Ground profileRadar altimeterGround heights along the track build the terrain profileFPGA
4 · Radar map matchGround profile, imaging radarProfile matched against the stored terrain map to correct driftFPGA
5 · Thermal scene checkThermal cameraTerrain picture checked against the reference sceneFPGA
6 · Update estimateRadar + thermal agreementAlong-track and cross-track update; course correction to the autopilot over signed serialARM

Barometer and airspeed also feed the core.

Sheet 03Flight simulationEngadin valley, Switzerland

GNSS outage rehearsal

Watch the flight simulation

The ShoalOS Simulator flies an authored waypoint route through the Engadin valley. UAV camera, imaging IR, radar and navigation-sensor panels run alongside the 3D scene, with an event log of each phase and GNSS state.

  1. Phase 01GNSS available

    Valley transit

    Follow the authored waypoints along the lake valley with GNSS available.

  2. Phase 02GNSS outage rehearsed

    Degraded environment

    Rehearse the GNSS outage while the route continues.

  3. Phase 03Outage continues

    Mountain passage

    Cross mountain terrain along the planned route.

  4. Phase 04GNSS recovered

    Observation orbit

    GNSS recovered; orbit the observation area.

Scripted animation from an authored route; the UAV is enlarged for visibility. Sensor panels are simulated or illustrative, not flight results. Imagery © Esri & contributors · Terrain © Mapzen / source providers · OpenGlobus.

Sheet 04The mission, step by stepSix steps

From flight plan through a GNSS outage

Six steps, from route planning on the ground to operator review.

01 / 06
  1. On the ground
  2. On the vehicle
  3. Across the squad
  4. With the operator
OutputReviewed flight plan
Planner

Start with a reviewed route

Select the destination and waypoints, calculate and review the route, then export the flight plan for the host autopilot.

Explore Nav Planner

Object tracking runs alongside navigation as a separate payload function. SquadSense and edge ATD/ATR are coming soon. The flight simulation is a scripted demonstration, not a flight result.

Fitted at build or retrofitted

Integrate the navigation payload

Works alongside your existing autopilot.