Today, we are excited to announce our support for the Avnet RASynBoard. The RASynBoard is a small form-factor, ultra-low power, edge AI/ML board, based on a Syntiant NDP120 Neural Decision Processor and a Renesas RA6M4 host MCU. The RASynBoard is ideal for always-on speech recognition, predictive maintenance, and industrial smart sensors. It is also ideal for battery-powered remote sensing applications due to its ultra-low power capabilities. The NDP120 subsystem features an on-board digital microphone, IMU motion sensor and SPI Flash memory.
We have partnered with Avnet, Syntiant, and Renesas to bring support for the RASynBoard to Edge Impulse. This will give users an efficient way to collect data, build models, and then deploy to the RASynBoard NDP120 Evaluation Kit.
With the help of Syntiant and Edge Impulse, Avnet has developed a RASynBoard FreeRTOS RA6M4 application that runs on the RASyBoard Evaluation Board called the Out-of-Box (OOB) application. The OOB application's goal is to "Provide a working example application that exercise the RASynBoard hardware and gives development teams a strong starting point for their own custom designs and ML training data." This fully documented project is implemented with a configuration file that allows the user to control the application behavior and to load new ML models developed in Edge Impulse without rebuilding any application code. Additionally, the application supports IoT connectivity to Avnet’s IoTConnect Cloud Solution on AWS or connecting directly to AWS to get your data into the cloud.
Model support for tasks such as motion classification, keyword spotting, and sound event detection.
To get started with your low-power embedded ML applications, visit the RASynBoard Getting Started Guide!