Developing Edge AI for Embedded Systems
Rekha R Nair, Tina Babu, Ebin P.M · 2025
Embedded systems have been revolutionized by the inclusion of artificial intelligence (AI), making them capable. This chapter gives a complete review of Edge AI for embedded systems with regard to its importance, software frameworks, hardware platforms, applications, and use cases. The significance of Edge AI in this paper demonstrates how real-time decision-making can be facilitated by Edge AI as well as the security and privacy aspects improved through local data processing. Moreover, the paper touches on the challenges experienced while implementing AI on embedded devices as well as underscores the importance of an effective algorithm for it. This book chapter examines AI software development and hardware platforms, focusing on popular frameworks such as TensorFlow Lite, ONNX Runtime, and OpenVINO, as well as hardware accelerators like NVIDIA Jetson motherboards, Google Coral USB memory sticks, and Intel Movidius VPUs. Additionally, this section explains their features, performance metrics, and how they fit into different types of signatures. This chapter also examines many examples of advanced AI used in various industries such as healthcare, automotive industry, smart cities, agriculture, or construction. It explains how edge intelligence can be used to perform tasks such as predictive maintenance or troubleshooting in areas with limited resources.