For a smart city project with hundreds to thousands of security cameras, it is almost impossible to have an adequate centralized AI server system due to the enormous amount of data to be processed. However, this problem can be solved with Edge AI Cameras as event data are processed and generated locally on devices.
With a traditional system, the camera must continuously transmit images to the server. The server will analyze the images and generate events and alerts which are then sent to the client's system. In contrast, Edge AI costs for data communication and bandwidth will be reduced as fewer data will be transmitted. The required compute power of the VA server will also significantly diminished.
With the AI on-device camera system, the AI event is handled right at the device, thus increasing real-time.
By analyzing images locally on edge devices instead of transmitting raw data to the server, the solution helps eliminate the need to send large and potentially sensitive data to the cloud, ensuring customer privacy as well as limiting cyber-attack threats.