PORRITTINC
( LAYTON, UT )
SPECIFICATION · EDGE THREAT DETECTION

Event cameras
and spiking
networks.

Neuromorphic image sensors report pixel-level change rather than frames. Spiking networks classify motion signature on-device, in darkness, at milliwatt power budgets.

← Back to Edge Threat Detection

DESIGN TARGET · NOT MEASURED PERFORMANCE

Figures on this page are engineering design targets and reference-architecture values, not measured results from operating hardware. They state what the design is being engineered to achieve. Verified performance data will replace them as systems are built and tested. Nothing here should be relied upon as a performance warranty.

Design basis
SENSOR

Change-driven pixels

Each pixel reports independently when its illumination changes beyond threshold. Static scenes generate almost no data; motion generates immediate output.

CLASSIFICATION

Motion signature

Classification distinguishes gait, vehicle, rotor, and animal signatures by temporal pattern rather than by appearance in a single frame.

CONDITIONS

Low light and high dynamic range

Per-pixel thresholding tolerates scenes that defeat frame cameras — headlights against darkness, shadow to sunlight transitions.

HUMAN ROLE

Detection only

The system detects and classifies. It does not identify individuals and it takes no action. A human receives the alert and decides what happens next.

Design targets
μs
Detection objective
mW
Node power objective
ON-DEVICE
Classification
HUMAN
On the trigger
Scope and boundaries

This is a detection and cueing capability. No facial recognition, no biometric identification, and no automated engagement of any kind. Output is an alert with context to a human operator. That boundary is a design constraint, not a configuration option.

Discuss the details.

Specifications evolve as the design matures. Contact us for current status and integration requirements.

CONTACT TIMOTHY → Back to Edge Threat Detection →