TechnologySignals and sensing

    Signals and sensing: which radio gives the accuracy your product needs

    Short answer

    We find out which radio gives the accuracy your product needs and how each unit behaves. In our own hardware lab we test Wi-Fi FTM/RTT, Wi-Fi CSI, UWB, BLE, LTE and mmWave radar, each against a written acceptance plan, then carry the option that passes into Rust firmware with per-device calibration, and sign what the device reports so nobody can fake it later.

    At a glance
    RadiosWi-Fi FTM/RTT · Wi-Fi CSI · UWB · BLE · LTE · mmWave
    Testedin our own hardware lab, against a written acceptance plan
    Modelspose and activity on mmWave point clouds
    EvidenceEd25519 on every message

    How a project runs

    Signals and sensing: from question to running system

    Your pathEach step ends in a written result
    1. 01Acceptance planaccuracy you need · how it is measured
    2. 02Lab testseach radio · against the plan
    3. 03Calibrationper device · live tuning
    4. 04FirmwareRust · OTA updates
    5. 05Signed evidenceEd25519 · witness quorum

    What we build

    Radio choiceWi-Fi RTT, UWB, BLE, LTE, mmWave radar and Wi-Fi CSI, each tested against a written acceptance plan, plus the hardware specification and procurement.
    Ranging accuracyWi-Fi FTM ranging on ESP32-C6 with a seven-step accuracy pipeline, live tuning over MQTT and per-device calibration, because every unit behaves a little differently.
    Machine learning on sensor dataPose and activity models trained on mmWave point clouds, and occupancy heatmaps.
    Phone signals and sensor fusionA mobile app that collects radio signals and sensor data on native Android and iOS, so a position is backed by more than one source and judged against how the device moved before.
    Signed evidenceEd25519 on every message and signed multi-party proofs with witness quorums, with relay, Sybil, replay and clock skew attacks mapped to a control before code was written.
    From signal to recordSigned evidence from devices through verifiers and a witness quorum to a blockchain record, so nobody can fake where a device was.

    Stack

    Technologies we use

    Radios and sensors
    Wi-Fi FTM/RTT, Wi-Fi CSI, UWB, BLE, LTE, mmWave radar, phone signal harvesting
    Firmware
    Rust 2024 on ESP32-C6, MQTT, over-the-air updates, per-device calibration
    Models
    Pose and activity models on mmWave point clouds, occupancy heatmaps
    Phones
    Native Android and iOS signal and sensor collection
    Security
    Ed25519 signing, witness quorums, threat modelling

    Proof

    Delivered work

    All work
    Case · Devices · Europe

    Devices that prove where they are

    Problem
    Signed proof of physical presence that nobody can fake.
    Built
    Rust firmware on ESP32-C6, radio lab research, signed proofs.
    RustESP32-C6Embedded firmwareWi-Fi FTM
    Firmware to chain, signed off sprint by sprint.
    Case · Location proof · Global

    Proof of location, built on sensor fusion

    Problem
    Prove where a device is without trusting the device.
    Built
    Signal-harvesting mobile apps, token contracts, Java back end.
    Proof of locationSensor fusionEthereumEOS
    MVP and token contracts that reward confirming nodes.

    FAQ

    Questions buyers ask

    Q01Which radio should we use for ranging?

    It depends on the accuracy your product needs and how your devices behave. We test the candidates (Wi-Fi FTM/RTT, UWB, BLE, LTE, mmWave radar, Wi-Fi CSI) against a written acceptance plan before the design goes into firmware.

    Q02Do you have hardware in-house?

    Yes. We test radios and sensors in our own hardware lab, each against a written acceptance plan.

    Q03Why calibrate every device?

    Radio ranging is noisy and every unit behaves a little differently, so each device gets its own calibration and can be tuned live.

    Q04Can a device fake where it was?

    Not if the evidence is signed and confirmed by others. We use Ed25519 on every message and witness quorums, and map relay, Sybil, replay and clock skew attacks to a control before writing code.

    Next step

    Working on signals and sensing?

    A 30 minute call with Mykhailo. You leave with a clear view of scope, risks and next step.

    Mykhailo Tiutin

    CTO, Vareger

    Mykhailo Tiutin, CTO, Vareger