Unmanned Systems Technology 015 | Martin UAV V-Bat | William Sachiti | Sonar Systems | USVs | Desert Aircraft DA150 EFI | SeaCat AUV/ROV | Gimbals

36 Focus | Sonar systems or less, hence the ultra-short baseline, which sends pulses to the towfish, which responds. The USBL then calculates the range and angle to the towfish. When using single-beam sidescan sonar in an AUV there are a number of considerations apart from the array size. The sonar has to be designed to provide high-resolution images from very shallow water to avoid multi-path interference. Onboard processing produces geo- coded sidescan imagery which is available for onboard computer-aided detection, classification and automatic target recognition, and the data needs to be compatible with post-mission analysis software packages. Covering a wide area, or swathe, often up to 200 m, means the resulting image has to be high quality for simultaneous search and classification. The images can be enhanced using a back-projection beam-forming technique that allows the focus to change for every pixel in the image, unlike most conventional sidescan sonars which have a fixed focus. Imagery from conventional sidescan sonar can be distorted by unpredictable movements of the AUV that come from the craft rolling, pitching or yawing, and this can leave gaps where there is not enough data to create the pixels. Even non-linear movements can be compensated for in software to give undistorted imagery with 100% coverage of the seabed. Capturing these images in high resolution in a sidescan array also needs highly linear element alignment, which cannot easily be achieved mechanically. Real-time array calibration is used to dynamically recalibrate each individual transmitter element several times a second to compensate for any dynamic strains causing array non-linearity. Multi-beam sonar Multi-beam sonar systems use beams from multiple transmitters with slightly different phases that reinforce in a particular direction. That allows more depth information to be extracted from a single ping, and the technique is widely used for hydrography applications to map the seabed in detail. These systems are usually fitted into the payload bay of the AUV or attached to the outside, which sometimes limits the size of the sonar array that can be used. For example, one 720 kHz multi-beam sonar provides real-time imagery of the underwater scene ahead of the AUV with a range of 100 m and a 120° field of view for obstacle detection, and a 20 º vertical view. The processing electronics in the receiver include a sensor that measures the speed of sound around the sonar head – knowing what that is allows the arrival angles of the various beams relative to the transducer head to be determined for more accurate measurements. The speed of the signals is also affected by the salinity of the water, which can change between the sonar head and the seabed or object, so knowing the sound speed profile from the transducer head to the seabed allows the sonar processor to calculate the range and correct for refraction of the acoustic beams. These are complex calculations that need to be combined with the amplitude, phase and timing of the sonar signals. The sonar head plus the multi-beam transmitters and receivers weighs just over 1 kg, and the unit is designed to be easily installed onto AUVs, connecting to the rest of the system via an Ethernet link. Systems are rated to different depths depending on the construction of the sonar head and the signal processing algorithms. Object detection is less depth- dependent than mapping, but multi-beam sonar systems are engineered to work down to 4000 m and even 6000 m for deep-water mapping. This is needed for oil and gas rigs, for example. August/September 2017 | Unmanned Systems Technology The results of using a multi-beam sonar towed by a USV to survey a wreck (Courtesy of Klein Marine Systems) Conventional sidescan sonar imagery can be distorted by movements of the AUV, leaving gaps where there is not enough data to create the pixels