Unmanned Systems Technology 025 | iXblue DriX I Maintenance I UGVs I IDEX 2019 I Planck Aero Shearwater I Sky Power hybrid system I Delph Dynamics RH4 I GCSs I StreetDrone Twizy I Oceanology Americas 2019

12 Researchers at the GRASP Lab at the University of Pennsylvania and the Tandon School of Engineering at New York University have developed an autonomous micro aerial vehicle (MAV) that can fly inside nuclear pressure vessels (writes Nick Flaherty). Accessing pressure vessels is a major challenge, as the pipes leading into it are less than 20 cm in diameter and the operating conditions can be adversely affected by condensation and fog. The 16 cm-diameter quadcopter weighs 236 g and uses two cameras for navigation and inspection, with onboard state estimation, control and mapping software running on a Qualcomm Snapdragon Flight module. It is powered by a two-cell, 7.4 V lithium-polymer battery. Researchers in Florida are working on a five-year project to develop autonomous ‘mother ships’ for UUVs and UAVs (writes Nick Flaherty). The $1.25 million project, at the Florida Atlantic University’s College of Engineering and Computer Science, will develop USVs that serve swarms of underwater vehicles such as the Bluefin UUV and UAVs used for coastal monitoring. The project combines multi-sensor perception, collision avoidance, simultaneous localisation and mapping (SLAM) as well as control algorithms to cope with bad weather. It will develop these components for a USV that will serve as a docking station for power and data transfer between the USV and the UUVs and UAVs. However, a small craft presents major challenges for processing data from the two cameras, as the baseline for stereo processing is small. The craft also needs to navigate in different illumination conditions, concurrently creating a map of the environment and re-planning its path to avoid any obstacles on the way to reaching its final mission goal. The front-facing stereo camera creates a dense map of the environment using two rectified images of 640 × 480 resolution at 15 Hz with 90 º horizontal FOV to determine the location of obstacles in the environment. To generate a high frame-rate map for real-time obstacle avoidance, the rectification process is split into two threads. Both images are concurrently rectified within 3 ms on the ARM processor in the Snapdragon. A block-matching algorithm produces a disparity map by searching for matches in the two images and providing distance information. Trade-offs between speed and accuracy can be explored by changing the block matcher filter size. Larger sizes generate less noise in disparity maps at a higher computational cost. The disparity information is then used to generate a 3D point cloud of the environment from a pinhole camera projection module. This allows the MAV to detect and avoid obstacles up to 25 cm in diameter, even with the relatively small stereo baseline. However, the team’s experiments showed that an onboard LED payload will be necessary to allow navigation in conditions with an illuminance lower than about 8 lux. UAV has a nuclear ability ‘Mother ship’ dock project Airborne vehicles Surface vehicles April/May 2019 | Unmanned Systems Technology The aim of the project is to develop USVs that can serve swarms of craft such as the Bluefin UUV (Courtesy of Florida Atlantic University)