BrainChip announced at Embedded World in Nuremberg Germany the launch of AkidaTag, a reference platform for smart sensing in a battery powered tag powered by Nordic Semiconductor.
Intelligent wearable and remote industrial sensors can deliver always-on sensing on a battery without relying on connection to a mobile, PC or cloud connection. Offering on-device adaptive learning allows personalization of the model while in use on-device, addressing the challenge of “one-size-fits-all” trained models and reduces the need to transmit data off the device for retraining on a GPU. OEMs can rapidly develop products based on the AkidaTag reference design using design partners.
The Nordic nRF5340 wireless SoC handles connectivity, sensor management, and application logic, while the BrainChip AKD1500 operates as a dedicated neuromorphic AI co-processor, delivering optimal energy efficiency. This provides a foundation for wearables and remote sensing device manufacturers to build devices that can interpret and communicate information effectively. AkidaTag features connectivity through Nordic Semiconductor’s nRF5340 Bluetooth Low Energy (LE) SoC on-device wireless communications. A BrainChip developed mobile application utilizes this connection to set-up configuration, load, and update models and firmware as well as receive diagnostics, logging, and alerts to the mobile device.
AkidaTag offers a blueprint and development kit with full design, hardware, firmware, and mechanicals that enable wearables makers and other manufacturers to build devices that can:
· Monitor biological signals for health and wellness applications while preserving privacy through fully on-device processing. Low power usage allows for days of monitoring vital health signals. · Detect anomalies in vibration and motion for classification in industrial equipment, enabling predictive maintenance · Voice wake-up commands for intelligent interfaces · Acoustic ambient environment detection and classification · Personalization of the device AI model using on-device Edge Learning & Self-learning through neuromorphic principles
“At BrainChip, we are committed to pushing the boundaries of what is possible at the edge,” said Sean Hehir, CEO of BrainChip. “By leveraging the robust ecosystem of Nordic Semiconductor, we are providing product designers with a blueprint for the future of wearables that’s always-on, privacy-first, and self-learning. This platform proves that high-performance AI and long battery life are no longer mutually exclusive.”
“The nRF5340 is the world’s first wireless SoC with two Arm® Cortex®-M33 processors, making it the perfect choice for complex IoT applications that require both high performance and extreme energy efficiency,” said Petter Myhre, Product Marketing Director at Nordic Semiconductor. “Seeing BrainChip use our technology to enable always-on neuromorphic AI demonstrates the potential of our wireless SoCs to drive the next wave of innovation in wearable and connected health markets.”
BrainChip will highlight the AkidaTag’s capabilities at Embedded World (Hall 5, booth 5-213) through a self-contained battery-powered “puck.” This will feature an integrated accelerometer for motion detection and a microphone for voice and vibration recognition. It can perform various functions, including keyword spotting, anomaly detection, and voice-activated wake-up. Control is through a mobile app that dynamically displays results and enables users to download activities for analysis.
Availability: The reference platform is licensed to accelerate development for OEMs and system integrators; including hardware schematics, firmware, and a companion mobile application. It will be available for evaluation in May 2026, with volume in Q3 2026.
