Benefits of Edge AI include faster insights, reduced network bandwidth consumption, increased privacy, and improved customer experience. While a portion of the edge AI potential is with more capable server-class hardware, there’s a massive opportunity to tap into the billions of embedded devices and systems in the physical world. Examples include sensors, cameras, gateways, controllers, robots, drones and cars.

 

Embedded development is complicated and the challenges are further amplified when trying to implement on-device AI. Models need to be optimized for limited hardware resources. Code integration requires close collaboration between AI and embedded developers but is cumbersome due to the monolithic nature of firmware. Edge AI necessarily increases the update cadence in the field as models are inevitably fine-tuned over time, requiring reintegration, retest and redeployment of the entire firmware image.

 

To effectively integrate and scale AI on embedded devices, developers would benefit from tools to build lightweight models that can be deployed as containers. Traditional containerization technologies like Docker can be too heavy for resource-constrained devices running embedded Linux, and are not an option for MCU-based devices. WebAssembly is set to redefine our approach to embedded development by bringing the benefits of Docker-like containerization to devices with as little as 256KB of memory.

 

In this webinar, you’ll learn about how Atym and Edge Impulse provide a complete solution for integrating, deploying, and managing AI models on resource-constrained devices at scale. You’ll learn about Edge Impulse’s comprehensive tool set for creating optimized AI models and Atym’s WebAssembly-based solution to enable and orchestrate containers for embedded devices.

 

Imagine the possibilities if you can implement the power of AI on the billions of resource-constrained devices in the physical world with the agility of containerization and modern CI/CD practices!