Issue 41 Unmanned Systems Technology December/January 2022 PteroDynamics X-P4 l Sense & avoid l 4Front Robotics Cricket l Autonomous transport l NWFC-1500 fuel cell l DroneX report l OceanScout I Composites I DSEI 2021 report

40 beamforming allows the system to steer the beam much more rapidly than a mechanical gimbal, while digital beamforming allows the user to form many beams and track multiple targets at different angles simultaneously. A phased-array radar sensor provides key advantages for SAA systems. A ground-based phased-array radar frees up payload requirements on the UAV, and a planar phased array can also provide a fast update rate and accurate angle and range information. A planar phased-array radar with a wide field of view angled directly at the sky can be used to create a 3D hemispherical bubble, in which the radar can detect targets and guarantee safe flight within it. The update rate for a phased-array radar that implements digital beamforming is the pulse rate of the system, which is of the order of milliseconds. As a result, such a system can provide high resolution in range, elevation angle, azimuth angle and time. The field of view of these systems covers the full extent of UAV flight paths, and can implement filtering methods to remove background reflections from the environment. This approach makes use of the technology enhancement for ADAS driver assistance systems and driverless cars. Frequency modulated continuous wave (FMCW) radar with digital signal processing enables the development of a 3D scanning radar that can use phased-array antennas to reduce the size and power consumption of the ground-based systems. One FMCW phased-array system uses 23 digital signal processing techniques to drive down the size, weight, power and cost of the radar’s design. The digitally steered system can track multiple targets simultaneously without the need for gimballed steering or mechanical scanning. In flight tests, the prototype transmitted 250 mW of power, which is sufficient to detect a small UAV with a radio cross- section (RCS) of 0.01 sq m at 250 miles or so. The radar sensor can detect larger aircraft at about 800 miles if the target has an RCS larger than 1 sq m. This is enabled by an algorithm called Recursive-RANSAC, which provides a robust and reliable tracking method that exploits radar measurements to track multiple intruders and distinguish between them in the presence of noisy, cluttered and missed measurements. The R-RANSAC algorithm, combined with a Kalman filter, provides an increased level of safety and integrity to an SAA system, and allows for an adequate time window for the collision avoidance logic to plan an evasive manoeuvre. The collision detection algorithm uses the geometric relationship between aircraft and estimates the time and distance at the closest point of approach to predict collision alerts that trigger the avoidance algorithm. Linear projections and geometric parameters can be computed efficiently, making it a tractable solution for multiple intruders. Moreover, prediction errors are negligible over short look-ahead time windows After R-RANSAC filtering, the time to closest point of approach (CPA) and the distance at CPA are computed to identify possible collisions. If a collision threat is detected, the intruder’s position December/January 2022 | Unmanned Systems Technology A 10 GHz FMCW phased-array radar controller for detecting UAVs (Courtesy of the University of Florida) A ground-based phased-array radar can provide sense & avoid capabilities for UAVs (Courtesy of the University of Florida)