Unmanned Systems Technology 018 | CES show report | ASV Global C-Cat 3 USV | Test centres | UUVs insight | Limbach L 275 EF | Lidar systems | Heliceo DroneBox | Composites

7 Platform one Subsea missions using synthetic aperture sonar (SAS) are difficult owing to the absence of real-time GNSS or tactical- grade IMU data (writes Rory Jackson). Kraken Robotics has therefore developed two proprietary motion compensation algorithms to enable real- time SAS missions – one of which has recently been integrated aboard a US Navy Remus 6000 AUV. The algorithms are embedded into Kraken’s Miniature Interferometric Synthetic Aperture Sonar (MINSAS) 120. The AUV’s position must be tracked to within one-sixteenth of a wavelength – which may run at about 300 kHz – for real-time beamforming to be achieved. The company offers two variants, both rated to a depth of 6000 m: the MINSAS 120 and MINSAS 60. The MINSAS 60 is a single 60 cm array, while the 120 is two 60 cm arrays assembled together. “All our MINSAS systems are modular,” said David Shea, vice-president of engineering at Kraken Robotics. “We build them up from receiver array blocks that are about 60 cm long each. And we have separate transmitters and receivers, so we can configure different systems in different ways. “The modularity allows us to reduce manufacturing costs and reduce end-users’ spares inventory and component costs.” The MINSAS 120 used by the US Navy’s AUV has four receiver arrays and two transmitter arrays. It uses Kraken’s RTSAS (real-time synthetic aperture sonar) pod, which contains an Intel i7 CPU and an Nvidia GeForce GTX 970M, along with the sonar’s electronics – power boards, amplifiers, transmitter electronics, field-programmable gate arrays (FPGAs), a fibre optic gyro and an inertial measurement unit (IMU). Shea added, “The CPU is mounted on a custom motherboard, and runs standard Linux. All our algorithms have been developed to be cross-platform, meaning we can run them in Windows or Linux, so we can develop and test different iterations of an algorithm very quickly before we port them over to Linux. “Linux has a much lower overhead than Windows for these kinds of applications. When you look at the data rates we’re dealing with, trying to stream in data from an FPGA over an Ethernet interface into this mobile computer, we need only minimal overhead to ensure the system is delivering the data without any delays.” That enables real-time beamforming through the RTSAS processor, and eliminates the need for post-processing (unless geo-referencing is required), which can add many hours to SAS missions, potentially generating 25 Mbit/s to be delivered over a Gigabit Ethernet link. Integrating a fibre optic gyro and IMU into the system avoids the need to integrate with an AUV’s navigation system, which for military unmanned vehicles is not always permissible. Real-time sonar realised Unmanned vessels Unmanned Systems Technology | February/March 2018 Kraken’s SAS has been integrated into a US Navy Remus AUV