Edge applications have tough, widely varying requirements. We’ll share lessons learned from three real-world applications in which Hailo’s deep learning processor is being used to implement deep learning. First, we’ll examine a video analytics use case, where multiple video streams are processed at the edge in real-time. Processing performance demands are substantial, and power consumption is a key constraint. Next, we’ll explore an industrial inspection application, where visual inspection informs the control of physical devices, resulting in a need for low latency. Finally, we’ll present a smart cities application, where network bandwidth is a key constraint, driving the need for processing near the sensor.