Powering IoT Developers with Edge AI: the Qualcomm RB3 Gen 2 Kit is Now Supported in the Edge Impulse Platform

This blog post was originally published at Qualcomm’s website. It is reprinted here with the permission of Qualcomm.

The Qualcomm RB3 Gen 2 Development Kit has been designed to help you develop high-performance IoT and edge AI applications. With powerful AI acceleration, pre-validated peripherals, and extensive software support, this kit enables every engineer to move quickly from prototype to production within a growing developer-friendly ecosystem. Now, with the integration of Edge Impulse,  you have even more flexibility to build, train, and deploy machine learning models on the Qualcomm RB3 Gen 2 platform.

High-performance hardware for IoT and AI at the edge

At the heart of the Qualcomm RB3 Gen 2 Kit is the Qualcomm Dragonwing™ QCS6490 / Qualcomm Dragonwing™ QCS5430 processor, delivering AI acceleration optimized for computer vision and edge inference. Whether you’re building for smart devices, robotics, industrial automation, or just any general-purpose edge AI applications for any data and model combination, this kit provides real-time superior AI processing with power efficiency in mind.

With Wi-Fi 6E connectivity for fast data transfer and expanded I/O support, the kit ensures seamless integration with external peripherals, sensors, or cameras. Built on the 96Boards open hardware standard, it allows you to customize and extend your setup with mezzanine cards to fit your application-specific needs.

A growing developer tooling ecosystem

With the addition of Edge Impulse platform, the Qualcomm RB3 Gen 2 Kit now offers an even broader set of tools for AI and IoT development. The Qualcomm developer ecosystem continues to expand, enabling faster prototyping and seamless AI model development and deployment, and allowing developers to move quickly from prototype to production.

Edge Impulse integration: AI model development made easier

Leveraging Edge Impulse Studio, you can now easily collect data, train, fine tune AI models, and deploy them on the Qualcomm RB3 Gen 2 Kit. Edge Impulse integrates with the Qualcomm AI Hub to ensure that your models are optimized for performance, power efficiency, and memory usage, allowing AI workloads to run smoothly on edge devices.

Through this integration, you will be able to:

• Use Edge Impulse Studio to train, and fine-tune AI models with real-world sensor data and deploy models as C++ libraries, Docker containers, or ready-to-flash binaries
• Leverage the Qualcomm Intelligent Multimedia Product SDK for advanced data pre-processing, post-processing, and inference

Get remote access with Qualcomm Device Cloud (QDC)

Don’t have a device to get started? With Qualcomm Device Cloud, you can start building and testing applications without needing physical hardware. This cloud-based Qualcomm Dragonwing™ QCS6490 gives you remote access to the physical hardware, so you can develop and validate real-time performances from anywhere. Combined with Edge Impulse’s MLOps workflows, you can deploy edge AI models seamlessly on our cloud hosted device accelerating your development cycle and making it easier to iterate before moving to real-world deployment.

Broad OS and software support

The kit supports multiple operating systems, including Linux (Yocto Project), Ubuntu, Android, and Windows, giving you the flexibility to choose the best platform for your applications.

Qualcomm Technologies provides a robust suite of developer tools, including:

• Qualcomm Visual Studio Code Extension IDE, ensuring seamless coding within VS Code
• Qualcomm Intelligent Multimedia Product SDK, enabling AI and multimedia processing
• Qualcomm Intelligent Robotics Product SDK providing specialized tools for robotics development
• Docker and containerization support, making AI model deployment scalable and efficient
• Over-the-air OTA updates with Foundries.io, ensuring continuous improvements and updates