Coilworks

Defence · NeuroCue Wearable

Wearable drone
detection.

We’re developing NeuroCue Wearable, a body-worn system for acoustic drone detection. Our next step is to evaluate the concept in representative conditions with pilot partners.

01 / HEAR02 / PROCESS03 / ACTBUFFERED MCUNEUROMORPHIC“TAKE COVER”“LOOK” DRONE BUZZRECORD → WAIT → BATCHBUFFERBATCHEVENT-DRIVEN SPIKESSPIKESSNNWEARABLE CUENTHREATHIGHLOWSECTOR2 O'CLOCKEXAMPLE CUESAME CUE · TWO PROCESSING PATHS DRONE
Illustrative pilot display

NeuroCue Wearable · Development goals

From ominous buzzing
to a clear next move.

You may hear a drone before you see it. But a buzz alone does not tell you what it is, where it is or what to do. NeuroCue Wearable is being developed to listen continuously on the body and turn that sound into one short, actionable cue — so the wearer does not have to stop and listen.

  1. 1.1 / Detect

    Hear it early.

    Recognise a drone's acoustic signature as soon as there is enough sound to go on, directly on the device.

  2. 1.2 / Classify

    Judge the threat.

    Classify what is heard, so the wearer knows whether it demands attention.

  3. 1.3 / Locate

    Know the sector.

    Indicate where the sound is coming from as a sector around the wearer, such as 2 o'clock.

  4. 1.4 / Act

    Hide or look.

    Turn it all into one clear prompt: take cover, or look towards the sector.

Development goals. Detection, threat classification and sector cues are in development and will be validated with pilot partners.

2 / Edge AI

Understand locally.
Share what matters.

Edge AI means interpreting sensor signals on the device. Instead of sending every raw signal away for analysis, the device can recognise a relevant pattern and provide an event cue.

2.1 / Recognise

Learn what to listen for.

Compact AI models help distinguish relevant sounds or vibration patterns from the surrounding activity.

2.2 / Process

Match the task to the hardware.

Where a buffered microcontroller records audio and processes it in batches, event-driven hardware responds to activity as it arrives. We combine it with conventional signal processing and compact models where useful.

2.3 / Inform

Turn signals into useful cues.

Provide an event, a condition trend or a directional cue where the sensing setup supports it — with response time and operating conditions considered together.

Pilot applications · Defence

Interested in a pilot?
Call us.

↗

We are looking for defence partners to take NeuroCue Wearable from the benchtop into realistic conditions.
Call us to discuss your use case and a first pilot.