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

83 correctly, avoiding the need to re-send data packets. All of this also has an influence on the choice of image sensor and the connection between the sensor and encoder. The resolution of the image sensor has not been a key factor in the past, as the unmanned platform could not send more than a high-resolution video stream, and a high-speed connection between the sensor and encoder was also not an issue. Now though, as more video can be carried, the sensor’s resolution can be higher, opening up the use of sensors with a resolution of 1080p60, 4K or 8K. Using ML in the video system also influences the sensor’s specification. No longer does it have to be a standard camera sensor with an array of red, green and blue pixels. A multi-spectral or infrared sensor can pick up additional information in other parts of the spectrum, and this information can be used by the ML system to increase the accuracy of identifying an object of interest and avoid triggering false positives, or identifying spurious objects. Ironically, that requires a faster connection to the encoder. Multi-spectral sensors deliver up to 10 times the data of a traditional camera sensor. There are several ways to handle this higher data rate, from a new standard from the MIPI consortium to 100 Gbit/s data links over fibre-optic links, and now even over a simple twisted copper pair cable. More efficient encoders also mean smaller power systems. DC-DC converters can be made smaller, saving weight and power in airborne systems. The current compression standard is the Advanced Video Codec (AVC), also called H.265 or the High Efficiency Video Codec (HEVC). This allows a 1080p30 video stream in a bandwidth of 1 Mbit/s down to 200 kbit/s at 15 fps with sophisticated additional tweaking of the video stream. That is half the bandwidth of the previous H.264 standard, and the progressive display provides a higher quality image for surveillance and analysis. All of this involves considerable trade- offs and tweaking, as most commercial system-on-chip encoders are developed and optimised for broadcast applications. The next generation of compression standard, now being developed, is called the Versatile Video Codec (VVC), H.266, which aims to support 8K ultra- high definition (UHD) video or 1080p60. The draft of potential technology for H.266 has been issued, and the final version is expected next year. Chip makers can then start working on a final version of video encoders for then and into 2021. In the meantime, AVC systems are constantly being tweaked to fit HD video over smaller links and in smaller boards with lower power consumption, to reduce the size, weight and power demand. These are being updated to HEVC, although that requires a slightly larger board, typically 4 mm longer than that for AVC. This similar footprint is an important factor, as it allows the higher compression rates to be added to a video system without increasing the size and weight of the components. The latest generation of video encoders also operate at lower voltages, typically 1.8 or 3 V. This allows for smaller DC- DC converters and enables system developers to shrink the overall size of the encoder system, with better power efficiency and improved resistance to the noise (called the ripple performance). However, these encoder chips have more complex power sequencing, as the power rails have to be activated in a specific order, so the design is more complex than previous boards. Video systems | Focus AVC systems are being tweaked to fit HD video over smaller links and in smaller boards, to reduce size, weight and power demand Unmanned Systems Technology | October/November 2019 This S-band bidirectional power amplifier operates from 2200 to 2500 MHz and draws just 0.65 A at 28 V to provide 2 W of transmission power (Courtesy of Triad RF)