FOR EVALUATIONDOC TORUS-AI-001REV 2026-07 CLASS UAGVIS / AI-ML enginePacketFive
TORUS AI/ML Engine

Intelligence at every tier of the sensing chain.

TORUS does not just detect, it decides. Machine learning runs where the data is born: on the node, at the gateway, and in command. Each tier turns raw signal into a smaller, more certain piece of intelligence, so what travels the network and reaches the operator is a classified track, not a raw feed or a bare alarm.

01 / WHY AI, NOT THRESHOLDS

A tripwire counts; an engine understands

Classify

Type, not just presence

Footfall, vehicle, or digging; voice, engine, or breaking glass. The system reports what it is, with a confidence, not merely that something happened.

Corroborate

Fuse across domains

A seismic cue plus an acoustic cue from a neighbouring node is a stronger track than either alone. Fusion is where confidence is earned.

Suppress

Fewer false alarms

Wind, rain, traffic, and wildlife are the real enemy of an unattended sensor. On-device models learn the background and hold their fire.

02 / THREE TIERS OF INTELLIGENCE

Edge, gateway, command

AI is distributed, not centralised. Every tier runs the inference its power and vantage allow, and passes a smaller, richer product to the next.

TIER 1 · EDGE / NODE

On-device TinyML

  • Seismic and acoustic classification on the ESP32-S3, quantized for a battery budget
  • Hardware wake-on-seismic, then an ML pass confirms and classifies before any radio use
  • Event vs background discrimination cuts the message rate and the power bill
  • Audio is classified at the node; raw audio never leaves it
ESP32-S3 · QUANTIZED TinyML
TIER 2 · GATEWAY / MEG

Accelerated edge inference

  • NVIDIA Jetson Orin runs thermal and visual inference on cued imagery from the masts
  • Detect, track, and classify at the gateway, so backhaul carries decisions, not video
  • Multi-node fusion across the local ring before anything leaves the site
JETSON ORIN · THERMAL / VISUAL / FUSION
TIER 3 · COMMAND / CCISRT

Correlation & tracks

  • Clusters corroborating events across nodes into a single track with aggregate confidence
  • Classification fusion, pattern-of-life, and false-alarm suppression across the whole field
  • Explainable rule-based correlator today; a learned model drops in behind the same interface
CCISRT · CORRELATION / TRACKING
03 / THE INTELLIGENCE PIPELINE

Raw signal in, decision out

SENSE

Raw signal

Geophone, microphone, thermal, LiDAR, RF at the node.

CLASSIFY

Edge TinyML

Node labels the event and its confidence; discards background.

FUSE

Gateway

Corroborate across the local ring; run visual inference on cue.

CORRELATE

Command

Form tracks across the field; suppress false alarms.

DECIDE

Operator

A classified track with location and confidence, ready to act on.

Each stage reduces data and raises certainty. A raw waveform becomes a labelled event, a set of events becomes a track, and a track becomes a decision, so the network carries kilobytes of intelligence rather than megabytes of feed.

04 / MODELS & METHOD

How the models are built and run

TierRuntimeApproach
EdgeESP32-S3, no acceleratorSmall quantized classifiers (TinyML) sized to RAM and the battery budget; wake-gated so inference runs only on a real cue
GatewayNVIDIA Jetson OrinAccelerated detection / tracking on thermal and visual streams; on-site so raw imagery stays local
CommandCCISRT serverRule-based correlation today (explainable, auditable); learned multi-target tracking and classification fusion drop in behind the same track interface
All tiersIsaac SimScenarios generate labelled data and validate detection and false-alarm rates before field trials

Explainable first. The command correlator ships as transparent rules so an operator can see why a track formed. Learned models are introduced behind that interface, with the rule-based path retained as a fallback and a sanity check.

05 / DATA GOVERNANCE & OPSEC

Classify at the source, keep raw data local

Privacy

Raw stays put

Audio and imagery are processed where captured; only compact labelled events cross the network.

Signature

Less on the wire

Sending decisions instead of feeds cuts bandwidth, power, and RF signature, which matters for a passive posture.

Assurance

Auditable inference

Track formation and command actions are logged, so an analyst can reconstruct why the system decided what it did.

06 / ROADMAP

From rules to learned fusion

NowNext
On-device seismic / acoustic classification at the nodeLarger vocabularies and per-site adaptation via OTA model updates
Rule-based correlation and track formation at commandLearned multi-target tracking and cross-domain classification fusion behind the same interface
Isaac Sim scenario data for validationActive learning from operator confirmations to cut false alarms over time
TRADEMARKS & THIRD-PARTY NOTICE

TORUS is an independent platform. Company names and product model numbers referenced in this document (including but not limited to Espressif ESP32-S3 and NVIDIA Jetson, Orin, and Isaac Sim) are used solely for engineering and bill-of-materials identification and imply no affiliation with or endorsement by those companies. Supply of any such third-party product to the designers, integrators, or evaluators of the TORUS platform remains at the sole discretion of the respective owning company, organisation, or legal entity. "FOR EVALUATION" is a document-handling marking only and denotes no government classification. All trademarks are the property of their respective owners.