BrainChip Holdings Ltd Unveils Edge AI Radar Platform for Defence and Autonomous Systems
Akida-powered solution delivers ultra-low power, real-time classification to improve decision-making in complex environments.
BrainChip Holdings Ltd (ASX: BRN, OTCQX: BRCHF, BCHPY), the world’s first commercial producer of ultra-low-power, neuromorphic AI technology, announced the launch of its Radar Reference Platform. This fully validated hardware and AI stack is designed to provide real-time object classification at the edge, solving the critical "identification gap" that limits traditional radar systems.
Solving the Identification Challenge
While standard radar effectively determines an object’s location and velocity, it often fails to identify exactly what that object is—a gap that can lead to false alarm fatigue or mission-critical failures. BrainChip’s new platform adds a sophisticated deep learning layer to traditional radar, utilising Micro-Doppler signatures to distinguish between targets with similar radar returns, such as a bird versus a drone.
From the Company
The Radar reference platform aligns with BrainChip’s overall solutions strategy.
"From drone countermeasures in the defence sector to non-invasive health monitoring in MedTech, the versatility of our Radar Reference Platform is transformative. We have moved beyond raw hardware to provide a complete, 'ready-to-deploy' technical stack that bridges the gap between raw data and actionable insights,” explained Sean Hehir, CEO of BrainChip. "While traditional radar has long been a staple of sensing, it has historically struggled with the 'identification gap'—the inability to distinguish between similar objects in complex environments. By bringing Akida’s™ neuromorphic intelligence to radar data, we are enabling a new era of classification at the edge. This platform doesn't just see movement; it understands exactly what it is seeing, providing the mission-critical intelligence required for modern defence and autonomous systems."
Neuromorphic Intelligence for Constrained Environments
The platform is engineered for environments where Size, Weight, Power, and Cost (SWaP-C) are paramount. By running natively on the Akida™ neuromorphic processor, the system delivers:
- On-Device Classification: Providing real-time inference without cloud dependency, ensuring operation in communications-denied environments
- Ultra-Low Power: Optimising for portable early warning systems and constrained autonomous platforms
- Environmental Resilience: Delivering robust performance through smoke, storms, and night conditions where traditional cameras often fail
Target Applications
The Radar Reference Platform is a versatile solution validated for five critical sectors:
- Defence & Tactical: Lightweight, deployable AI for real-time threat classification and "Identify Friend or Foe" (IFF) capabilities
- Drone Countermeasures: Specific detection of propeller signatures to trigger precision responses.
- Health & Biosignals: Contactless fall detection, activity monitoring, and gesture recognition without invasive video recording.
- Marine & Autonomous Platforms: High-efficiency obstacle detection and navigation for unmanned surface and aerial vehicles.
- Robots and Autonomous Vehicles: Environmental navigation for robots and delivery vehicles in adverse conditions.
A Ready-to-Deploy Technical Stack
To accelerate time-to-market, BrainChip provides a complete, integrated stack:
- Hardware: BrainChip AKD1500 co-processor paired with an Asahi Kasei FMCW Radar Module
- Software: A pre-integrated Micro-Doppler classification model and a real-time dashboard for visualising Range-Doppler and Micro-Doppler plots.
Live Walkthrough Webinar
BrainChip will host a live technical deep dive of the Radar Reference Platform at 8 a.m. PT on April 20. Principal Product Manager Nick Markovsky and Machine Learning Engineer Amir Naderi will demonstrate the platform’s architecture and live object classification capabilities.
