Unmanned Systems Technology 010 | nuTonomy driverless taxi | Embedded computing | HFE International marine powertrain | Space vehicles | Performance monitoring | Commercial UAV Show Asia report

10 Platform one Qinetiq is to provide the UK component for a multinational demonstration of cooperating unmanned and autonomous systems under contract to the Defence Science and Technology Laboratory (DSTL) (writes Peter Donaldson). The Hell Bay 4 demonstration is to take place off the coast of Scotland as part of the Royal Navy’s Unmanned Warrior event in October, and will involve UAVs, USVs and UUVs working together in squads on mine countermeasures missions and supplied by companies including SeeByte, Blue Bear Systems Research and Autonomous Surface Vehicles (ASV). The first phase of the DSTL-led Maritime Autonomy Framework initiative established software and architecture to test UUVs, while the second focused on integrating a USV into a cooperative squad. Under the third phase, a UAV is to be integrated to provide situational awareness and a communications relay, and will bring together vehicles operated by the Navy’s autonomous systems trials team plus the ASV C-Worker 5 USV and the Blue Bear Blackstart UAV. The wider Unmanned Warrior event is to take place around the UK coast, and will involve 40 organisations, including the US Navy and NATO’s Centre for Maritime Research and Experimentation, along with 50 vehicles. In addition to mine countermeasures, activities will focus on anti-submarine warfare, ISTAR and C2, as well as hydrographic and geospatial intelligence. All systems go for Warrior Cooperative systems A team from Professor Vijay Kumar’s lab at the University of Pennsylvania has shown that quadcopters can perform a range of aerobatics stunts, including flying through small open windows, using only onboard sensing, processing and path planning (writes Peter Donaldson). However, this requires obstacle positions in advance. Earlier demonstrations relied on external cameras, position tracking, path planning and data links. The 250 g quadcopter used for the demonstration was equipped with an IMU, a Qualcomm Snapdragon processor and a Hexagon DSP – all of it smartphone-grade hardware. The demonstration, in which the quadcopter rapidly and repeatedly assumed a vertical bank angle to pass through a window too narrow for it to pass through in horizontal flight, is the result of six years of work aimed at enabling useful autonomous operations in realistic, obstacle-filled environments. The craft has a front-facing stereo camera for ‘dense mapping’, which is soon to be integrated into a real-time planning and control framework. Indicating that the team has solved the problem of state estimation, control and planning, the UAV gained momentum and changed attitude in one smooth motion to pass through the window rather than dividing the manoeuvre into separate phases. It obtains its position and velocity estimate using camera data and closes the feedback loop at 500 Hz, enabling it to manoeuvre relatively hard, pulling 1.5 g , flying at 4.5 m/s and making 90° changes in pitch and roll. The size of gap it can go through depends on knowing its position and that of the gap in the window with a very high degree of certainty. The next step is to get it working outside the laboratory environment, so the team is developing real-time mapping strategies for obstacle detection and dynamic planning. Leap forward with aerobatics Aerial autonomy October/November 2016 | Unmanned Systems Technology The Blackstart UAV will be part of a demonstration involving unmanned air, sea and submersible craft