PYLON Patent pending Rev A prototype

The last mile between soldier sensors and the tactical network.

PYLON is a low-SWaP body-worn node that aggregates body-worn Bluetooth sensor data, processes and encrypts it at the edge, and bridges it to the tactical network over a low-probability-of-intercept LoRa mesh, delivering an individual position for every operator below the squad leader. One end-user device per squad renders the full picture in TAK.

PYLON is an underlay, not a replacement. It complements program-of-record radios and leader end-user devices, and every inter-node link is an open standard.

PYLON Rev A enclosure, engineering render

Engineering render of the Rev A enclosure.

PYLON owns the critical first hops: ingest, process, encrypt, compress.

Data arrives at the tactical network ready to transport, not as a raw dump. Everything downstream of PYLON is infrastructure the customer already owns.

Secure BLE-to-TAK gateway

Aggregates body-worn Bluetooth sensors, encrypts at the application layer, and formats output as Cursor-on-Target for native ATAK ingest. No gateway server, no custom integration.

Edge processing and compression

On-device processing filters, prioritizes, and compresses raw sensor streams before the first network hop. The mesh carries events, not raw streams. Design target: at least 90 percent bandwidth reduction.

Individual position for every soldier

Multi-constellation GNSS with IMU dead reckoning from the last fix through GNSS denial. Enables fires deconfliction, casualty evacuation, fratricide prevention, and UAS deconfliction below the squad leader.

Current design and production target, stated separately.

The left column is what exists on the bench today at Technology Readiness Level 4. The right column is the funded-effort target. We keep the two apart on purpose.

ParameterCurrent design (TRL 4)Production target (funded effort)
Form factorMOLLE-mountable body-worn node with sealed enclosureRuggedized overmold, same envelope
Weight94 g with battery (design weight)85 g
EnvironmentalSealed enclosure selected for ingress protectionMIL-STD-810H subset qualified: drop, ingress, thermal
PowerSingle-cell lithium-polymer battery, USB-C recharge; multi-day mesh-node design budget, under one watt peakValidated 96 hours or more at full capability, 144 hours or more in cache-and-forward
ComputeArm Cortex-M4F Bluetooth 5 SoC in a certified module with integrated PA/LNASame
Mesh radioSub-GHz LoRa transceiver, chirp spread spectrum, 902-928 MHz ISM bandFirmware-driven channel hopping across the configured sub-GHz plan
Transmit behaviorShort, infrequent bursts at a low duty cycleDuty-cycled per emission-control mode
RF signatureLow transmit power and spread-spectrum modulation, designed to sit below narrowband detection thresholdsDetectability characterization in government experimentation
Onboard sensorsSix-axis IMU, multi-constellation GNSS, environmental (temperature, humidity, pressure)Same
Sensor aggregationBluetooth 5 body-worn sensors: physiological, environmental, missionExpanded peripheral profiles
Individual PLIGNSS 1.5 m CEP nominal per chipset specification; IMU dead reckoning from the last fix through GNSS denialDead-reckoning performance validated under denial scenarios
Data outputEvent-driven Cursor-on-Target on open standards; CivTAK native, ATAK-compatible; hands off to existing radios unchangedTAK Server integration; bandwidth-reduction target instrumented
SecurityApplication-layer encryptionFrame authentication; cybersecurity documentation package
Emission controlSealed, glove-operable hardware kill: RF power gate at the rail; no software or remote command can reactivate the radiosFour-mode selector: full operations, reduced signature, cache-and-forward, radio silent
CostAll-COTS bill of materials; unit pricing on requestHeld at production quantities

Specifications describe the Rev A design under bring-up and are subject to change through qualification. Full data sheet available on request.

“If every soldier is transmitting, aren’t you painting the formation?”

PYLON is designed from the ground up for contested electromagnetic environments. The LoRa link runs at a small fraction of a tactical radio's transmit power, in short bursts at a low duty cycle. Chirp spread spectrum pushes power spectral density below narrowband detection thresholds, and the self-healing mesh keeps each hop short. When the situation calls for silence, the operator gets it in hardware.

Mode 1
Full capability

Bluetooth and LoRa mesh active. All sensors reporting. Training and permissive environments.

Mode 2
Reduced signature

Bluetooth at minimum power. LoRa disabled at the hardware level, not in software. Contested environments.

Mode 3
Cache and forward

Bluetooth receive only. LoRa off. Sensor data logged to non-volatile memory and burst-transmitted when the situation permits.

Mode 4
Radio silent

All radios off through a hardware kill switch that physically opens the RF path. Zero emissions. No software or remote command can reactivate.

As built today: sealed, glove-operable hardware kill at the rail. The four-mode selector is the funded design.

A sensor-to-network bridge, not a tactical radio.

Tactical radios, soldier hubs, and end-user devices each do one job well and each stop at the leader. PYLON sits underneath all of them and extends the network to the soldiers who carry none of them. It hands existing radios pre-compressed Cursor-on-Target, it gives soldier hubs a peripheral their architecture already expects, and it gives the leader's end-user device a picture of the whole squad instead of a picture of the leader.

PYLON is
  • The aggregation tier beneath the radio, the hub, and the display
  • An individual position and sensor bridge for every soldier, priced per soldier rather than per leader
  • Native Cursor-on-Target output into ATAK, CivTAK, and TAK Server
  • Hardware-enforced emission control with data preserved through radio silence
  • Open standards on every inter-node link, vendor-neutral toward the radio and the hub
PYLON is not
  • A tactical radio, and it does not replace one
  • A soldier computer, a display, or a battle management application
  • A voice system
  • A replacement for the leader's end-user device
  • Dependent on a vehicle, a tower, a satellite, or a server to form its network

TRL 4. Rev A prototypes in hand and operating as a multi-node mesh. Rev B revision in design.

All development to date has been company-funded ahead of any government dollar, and all intellectual property is founder-developed and company-owned.

Mesh and PLI validation
End-to-end chain demonstrated on development hardware: GNSS position into a four-node self-healing LoRa mesh into CivTAK as Cursor-on-Target, with per-operator icons on the tactical map.
Circuit board
Custom Rev A board designed with a U.S. electrical-engineering partner: Bluetooth 5 SoC with integrated PA/LNA, dedicated LoRa transceiver, multi-constellation GNSS, six-axis IMU and environmental sensor, non-volatile flash. All-COTS bill of materials validated at multiple quantity tiers. Twelve assembled units delivered August 2026.
Firmware and interop
Custom Bluetooth beacon and selective gateway scanner firmware on development hardware; Cursor-on-Target generation validated against CivTAK. All twelve Rev A units pass self-test on the sensor, GNSS, and Bluetooth subsystems. September 2026: four Rev A units running as a Bluetooth mesh with relay, nodes on battery, plotted live in CivTAK; sealed hardware kill and Bluetooth-only reduced-signature mode bench-demonstrated.
Companion application
Android application framework for sensor onboarding and node configuration, built on the Nordic Bluetooth library.
Enclosure and SWaP
Three MIL-STD-810H enclosure options analyzed (aluminum, glass-filled nylon, polycarbonate with TPU overmold); overmold selected for production. Additively manufactured nylon enclosures for the Rev A prototypes in production.
Intellectual property
U.S. patent applications pending on the core architecture, including emission control and the mesh waveform. All IP founder-developed and company-owned.
Next
Rev B board revision (antenna and RF front end); company-funded squad-scale validation, Q4 2026; TRL 4 to TRL 6 with government funding through a nine-month maturation aligned to sponsor experimentation.

The same node serves every operator who works where the network fails.

Law enforcement tactical teams in transit tunnels, wildland hand crews beyond repeater coverage, incident commanders at a mass-casualty scene, underground mine supervisors, and expedition crews far from any tower all need the same three things: where their people are, what condition they are in, and a way to stay connected without infrastructure. PYLON's hardware, firmware, and open output format are designed to serve each of them.