AI at the Edge

Chapter 13. Use Case: Consumer Products

Edge machine learning is used in consumer electronics and products to enable devices to make decisions based on data without sending that data to the cloud. This can save time and bandwidth and can also be used when data is sensitive and needs to be kept private. Edge machine learning can also be used for consumer-focused tasks like facial recognition, object detection, voice recognition, and sensor classification. By analyzing and recognizing patterns in the consumer data being ingested on the device before sending it to the cloud for further processing, products can quickly adapt to the user’s needs: show the desired product usage, provide customized alerts to the user about the product, and more.

By using edge AI, consumer products can integrate with and leverage the data of onboard sensors for an almost unlimited amount of use cases. For example, a bike can analyze the rider’s surrounding environment for traffic information and environmental data that can affect ride quality, and a smart fridge can automatically detect when a product has almost been used up and add the product to a purchase list. In this chapter, we will brainstorm various approaches to using edge AI for consumer products, their associated sensor and device configurations, and a deep-dive tutorial into our selected approach and use case solution.

Problem Exploration

Many consumer technology products are already constantly connected to the internet, such as smart home ...