The basic components of a typical edge artificial intelligence model include: hardware and software for capturing sensor data, software used for training models in different application scenarios, and applications running artificial intelligence models on IoT devices. The microservices software running on the edge device is responsible for launching the artificial intelligence package on the edge device according to the user's requirements. Within the edge device, feature selection and feature transformations determined during the training phase are used. These models can be customized to the right combination of features that can be extended to include aggregation and engineering features.
Smart edge devices are deployed in battery-powered applications with narrow bandwidth and intermittent network connections. Edge device manufacturers are therefore building sensors that have integrated processing and storage capabilities, using widely used low-speed communication protocols such as BLE, Lora, and NB-IoT, which are small in space and low in power consumption.