Unmanned Systems Technology 028 | ecoSUB Robotics AUVs I ECUs focus I Space vehicles insight I AMZ Driverless gotthard I InterDrone 2019 report I ATI WAM 167-BB I Video systems focus I Aerdron HL4 Herculift

82 A major change is taking place in how video for unmanned systems is managed. Video compression technology is evolving so that the bandwidth required for HD video can be further reduced compared with current systems, which will allow high-quality video over narrower bandwidth links and support higher resolution streams over current links. This is appropriate for many applications, especially unmanned ground vehicles, for autonomous control as well as through a remote operator. At the same time though there remain considerable technical challenges for video. One is the latency of the links and the use of satcom. For current satellite links, the latency is 600 ms each way, giving a total delay of 1.2 s for controlling a gimbal for example. That makes it very difficult to position the camera accurately to track an object of interest. There are also internal issues with gimbal systems in unmanned aircraft. The 1.5 Mbit/s bandwidth across the gimbal’s slip rings limits the quality of the video. For example, systems using HD video at a frame size of 720 pixels and a frame rate of 60 Hz (called 720p60) can be delivered over the slip rings reasonably easily. Video with frames 1080 pixels wide at 30 Hz (1080p30) are still possible, but system makers want to deliver 1080p60 video over the slip rings. That is why many of the sensor makers are trying to embed the video compression inside the turret that houses the camera. Both of these issues are leading to a dramatic increase in the use of machine learning (ML) and AI approaches to avoid having to send the whole video. These algorithms can identify areas of interest such as people or vehicles and then just send the relevant areas of the video frames. This can significantly reduce the bandwidth required for a link, allowing satellite links of 600 kbit/s to be used and maintain a high quality of video stream or even send a series of still images. That means AI processing chips are also being used for video systems alongside (or instead of) the traditional video compression system. Field- programmable gate arrays (FPGAs) have been widely used to implement proprietary algorithms to improve the quality of the video and how it is transmitted. This can include proprietary forward error-correction implementations that add correction codes to the data stream to ensure the data is received Nick Flaherty provides an update on developments in video for unmanned systems, and the challenges and issues that remain A moving picture October/November 2019 | Unmanned Systems Technology The STAMP-VMD camera gimbal is a complete video turret for a UAV such as the Orbiter, flying at an altitude of 500 to 1000 m (Courtesy of Controp)