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

74 and accurately gauge its distance to determine when the DroneCatcher was within net-shooting range,” says Rath. “That carried over nicely to the Kotug project, to prevent the UAV colliding with the ship, by enabling it to detect range and to inch closer.” “The EO and Lidar are coupled in a gimbal, which we designed and built, along with the software and electronics. We also selected the sensors we thought were best, before writing our own software to fuse their data and output it to the user.” Delft Dynamics also put a lot of effort into selecting navigation sensors for the task, as the ship’s electronics and metallic hull could disrupt the UAV’s GNSS receiver. That meant using inertial and barometric sensors calibrated well enough to provide for dead reckoning accurate to a few centimetres over 20-100 m. “In another project, we had to approach a large chimney made from steel, which gave us something of a benchmark to use with Kotug’s operation,” Rath notes. “At around 10 m, the magnetometer – generally an important sensor for GNSS orientation – experienced problems. If that fails, left, right, forwards and back can rotate around the z axis and you lose the stability of your controls.” To that end, the RH4 ferrying the heaving line has had its magnetometer removed, relying fully on its accelerometer and gyroscope to remain stable while airborne. The two companies have found that works well. Once the UAV reaches the Panama chock, it must somehow attach the line to or near the chock. The companies are trialling a few options for doing this. The most promising method, Smoor comments, involves the use of magnets. The heaving line would have a magnet attached to the upper end, initially ‘mated’ to another magnet on the UAV, and would be attached onto the ship’s deck via a stronger magnet next to the Panama chock. Alternatively, a latching mechanism could be used to secure the line to the UAV, as the RH4 could need a magnet as heavy as 1 kg to keep the line stable at full length. The combination of an ejection mechanism and a hook is also being explored, to help fire the line through the Panama chock and hook onto it, or onto some other capture system. That is lighter but mechanically more complex, and thus more prone to malfunction. It also requires more precise angles for reliable aiming, whereas magnets are less demanding in that respect. As Rath points out, “The Panama chock is about 25-50 cm in size, so it’s very small to have something like a weighted line fired through it. It’s similar in size to a lot of consumer drones in fact, but we need to shoot through the middle of it with greater accuracy than you’d need for a net-gun, so we need further trialling and optimisation than for the DroneCatcher.” Once the line is sufficiently secured to the Panama chock, whether by magnet or hook, the UAV will fly back towards the tug. An acoustic signal (or a flashing light, particularly at night) will be produced from the UAV to indicate it is safe for the crew to approach and retrieve the heaving line from the Panama chock, and pull it up by hand. They will then tie it to a nearby winch, April/May 2019 | Unmanned Systems Technology Delft Dynamics uses a combination of cement-like adhesives, custom-designed seals, and closed electric motors to prevent the RH4 being affected by water ingress as it repeatedly crosses the water between tug and ship