Level 42 AI · DIU Low-Cost ISR (LC-ISR) Challenge

Passive sensing for the contested edge

We fingerprint how a place behaves and flag the instant it changes. Left of boom.

See how it works

The problem

Legacy ISR fights right of boom

Legacy ISR is a contour map plus a signature library: it knows where things are and what it has already seen. Like early genome annotation, it recognizes only what its frameworks were built to see — not what has changed or is changing.

Left of boom

Sense, cue and act before contact

  • Asks “How is this place changing?”
  • Flags threats it has never seen
  • Expeditionary: no runway, minimal crew
  • Attritable, affordable mass; zero emissions
  • Cues the kill web before visual emergence

Right of boom

React after the visual. Too late.

  • Asks only “What is here?”
  • Blind to anything never cataloged
  • Tethered to airfields, heavy on crews
  • Too exquisite to risk, too few to mass
  • Kill chain starts only after a visual
  1. Sense
  2. Understand
  3. Decide
  4. Act
  5. Assess

Urgency: every day without affordable passive sensing is another day adversaries move inside the gaps between expensive ISR looks.

Core insight

Stop asking “what does it look like?” Ask “how did the field break?”

Threats adapt signatures. Physics still disturbs pressure, vibration, electric potential and timing.

Coastline rendered as a terrain model with acoustic rings spreading from a vessel through the ambient field
  1. 01

    Ambient field before contact

    Traffic, wind, HVAC, surf, rotor wash, engine vibration and human activity form a background pattern with stable local structure.

  2. 02

    Target enters

    Anything with mass, motion, frequency, wavelength or electric-field contrast perturbs that pattern.

  3. 03

    AI estimates the break

    Group-symmetry, topology and sequence models turn disturbance geometry into bearing, localization, class and activity state.

  • No active emission.
  • No prior acoustic signature.
  • No dependency on a perfect visual look.

The paradigm: V!broKinetic Topology Fingerprinting

Every target breaks the ambient symmetry

Level 42 AI’s Anthropophony™ platform turns ambient battlefield disturbance into early cueing, custody and downstream action.

Coastal operating area at dusk: a drone deploys passive sensor nodes while rings show disturbances spreading through the ambient field
Ambient field · AnthropophonySymmetry break → passive bearing + localization

  • Biomimetic by design

    Like a platypus in muddy water or a tiger beetle sprinting blind, we sense by vibration, pressure and timing, spending a small data budget only on change.

  • From contour map to fingerprint

    Contour maps show where things are: position, elevation, slope. We fingerprint how a place behaves: topology, frequency and vibration over time.

  • Deviation becomes the signal

    Anything with mass and motion breaks the ambient symmetry in pressure, vibration and electric potential. The break is actionable.

  • Nothing hides from physics

    No signature library. No emission. Even threats never cataloged, never seen before, cannot hide from physics.

Spectrum, beyond light

Seven orders of magnitude. Seven beyond human perception.

40 HzAudible

Drag across the scale. The shaded band is what human ears can hear; the rest is where the payload keeps listening.

Beyond human perceptionHuman senses cover a narrow slice of the field. The payload hears, feels and sees the rest.
  1. Sense firstThe passive stack holds a persistent stare and catches the break.
  2. Act firstBearing and class cue the kill web before visual emergence.
  3. Win the decision cycleSeen first, understood first: the adversary reacts to you.

Sensor stack

A carrier-agnostic passive ISR payload for the transparent battlefield

Level 42 AI integrates vibroacoustic sensing, electric-potential sensing, bowtie aperture geometry and edge AI.

Exploded view of the ruggedized passive ISR payload: transducer disc, antenna plates, sensor boards, edge AI board and battery

Design principle: cheap enough to proliferate, quiet enough to survive, open enough to integrate.

The asset

The asset is not the sensor. It’s the data corpus and the algorithms.

A carrier-agnostic passive ISR payload for the transparent battlefield.

Fast days to weeks

Acute infectious disease

Slow months to years

Chronic non-communicable disease

Slowest years to decades

Maternal–fetal development and degenerative disease

L42 ISR drone with red rotors Passive ground sensor node in two housings Vibroacoustic sensor pod on a mounting rail

The aerial platform: L42 ISR drone

Built for the austere edge, not the airfield

In integration · design targets
3–24+

hours endurance

Medium (3–12+ h) and long (12–24+ h) profiles from one modular power system

<1 hr

offload to launch

Planned, offloaded and launched by a minimal crew

IP67

ingress protection

Operates through dust, water and shock

0

runways required

Austere launch, recovery, refuel and refit

Offload-to-launch time (hours)

Government baseline3 h
L42 target<1 h
Wireframe of the L42 ISR drone titanium airframe with four ducted rotors

One airframe, both categories

  • Minimal crew for launch, recovery and maintenance
  • No hangar, runway or fixed infrastructure
  • Flight-hour evidence delivered before fly-off

The payload: multiphasic sensor stack

A passive stack that cues the kill web

Audra™ broad-bandwidth transducer

Vibroacoustic: infrasonic, audible sound to ultrasonic environmental signatures

Demonstrated

Far-field capacitive electric potential sensors

Electric: passive E-field disturbance, no emission

Demonstrated

High-sensitivity bowtie antenna stack

RF: broadband anomaly detection

In integration

SWaP-optimized: multi-int, decision-quality data within the size, weight and power budget of inexpensive airframes.

Passive sensor-to-shooter cueing

  1. Find

    Low-power passive stack holds persistent stare

  2. Fix

    Neurosymbolic AI confirms the symmetry break and bearing

  3. Cue

    EO/IR, SAR or SIGINT wake only now; hand-off to shooters

The neurosymbolic AI engine: Anthropophony platform

Five orthogonal layers. One decision chain.

Classify

H1Topology

Persistent homology

Classify scene type and target class

Demonstrated

Confirm

H2GF Capacitive

Vibroacoustic pressure × electric cross-terms

Correlated activity confirmation

Demonstrated

Anticipate

H3Koopman HMM

Operator theory

Track activity sequences, anticipate state

In development

Pre-cue

H4Hamiltonian Bifurcation

Phase-space transitions

Pre-event cueing

Demonstrated

Locate

H5Symmetry Breaking

Peter-Weyl decomposition

Immediate passive bearing and localization

Demonstrated
Like the gaps in Mendeleev’s table, the structure holds space for the unknown, flagging a threat before it has a name.

The V!BEC2 interface

Superhuman passive sensing for every contest

Demonstrable today
Zero learning curve: familiar workflow, built for the warfighter, not the specialist.

  • Hearthe inaudible

  • Feelthe imperceptible

  • Seethe invisible

Contested, DDIL & the transparent battlefield

Silent by design: nothing to detect, jam or spoof

0

RF or active radar emissions

On a transparent battlefield, if you emit, you die. We classify and take bearing from disturbance alone. We never emit.

Exploded passive sensor node

Passive survivability Demonstrated

Detection, classification and bearing with zero RF or radar emissions: no signature for the adversary to hunt.

GNSS-denied navigation In development

Orientation and terrain sensing from the local mathematical topology of the ambient field, with no satellite dependency.

Open, secure dissemination Planned

Sensor data over Government-directed radios and waveforms, LOS and BLOS, via our C2 integration TDP (interface docs + SDK/API).

Universal domain execution

One passive architecture across land, littoral and maritime seams

Vehicle sensed on entry
Ambient field
Urban anthropophony: traffic, HVAC
Symmetry broken
Horizontal translational symmetry

Vehicles sensed on entry

Autonomy & Modular Open Systems Approach

Open by architecture, autonomous at the edge

  1. Sense

    Vibroacoustic · E-field · RF + third-party EO/IR, SAR

  2. Decide at the edge

    Agentic Neurosymbolic AI: H1–H5 automated detection and alerting

  3. Open interfaces

    Documented APIs, SDK and C2 integration TDP

  4. Integrate

    Project AVO · Gov & third-party C2/UI · fires · intel

  • Reduced Soldier workload

    Edge-computed agentic AI runs detection workflows and alerts automatically, cutting cognitive load on tactical operators.

  • Government-accessible interfaces

    Payload, data, C2 and radio interfaces are open and documented; only H1–H5 model internals are proprietary.

  • Built for Project AVO

    MOSA alignment designed for the Autonomous Vehicle Orchestrator and collaborative autonomy.

Field validation

58th MDW Warfighter and DTRA field tested

A service member places a handheld sensor on a patient lying in the cargo hold of a C-130 air ambulance, with readings on a tablet
59th MDW Air Ambulance C-130 Hercules
A pilot in a flight suit seated in an ejection seat test rig with a laptop recording beside him
Close view of a pilot strapped into a Martin-Baker ejection seat during a measurement session
A Royal Dutch Air Force pilot in an ejection seat smiles while a screen behind her shows a body signal map
Royal Dutch Air Force / Martin-Baker F18

Capability maturity map

What’s real today, what’s integrating, what’s next

  1. Today

    Demonstrated

    Demonstrable today

    • Vibrome® nCoder
    • Moto Smart Scanner integration
    • H1 Topology + H5 Symmetry Breaking algorithms

  2. Fly-offs · FY27

    In integration

    In development / integration

    • L42 ISR robodrone flight platform
    • Far-field array + bowtie antenna stack
    • H2 / H4 fusion layers
    • GNSS-denied navigation · C2 TDP
    • Project AVO collaborative autonomy

  3. Beyond FY27

    Projected

    Projected future capability

    • V!bro Living Contour Mapping of the operating area
    • Multi-domain drone swarms
    • Synchronized H3 Koopman HMM anticipation across thousands of square miles

Sense first. Understand first. Act first.

Let’s field left-of-boom ISR together.

info@level42.ai +1 650 515 6112