Uncrewed Systems Technology 043 l Auve Tech Iseauto taxi l Charging focus l Advanced Navigation Hydrus l UGVs insight l MVVS 116 l Windracers ULTRA l CES 2022 show report l ECUs focus I Distant Imagery

40 Focus | Battery charging an open software development kit, and all the comms protocols and tools are open source. Driverless cars Connector systems for electric ground vehicles face more challenges than UAVs for charging larger battery packs. One approach uses connectors that can be lowered from the underneath of a vehicle to connect to a pad with a matrix of connector areas on the ground. This can deliver up to 22 kW of AC power and 50 kW of DC charging with 99% efficiency. At these higher charging levels, the efficiency is a key figure to reduce the amount of energy wasted as heat during the transfer. Although taking up just 44 mm in the vehicle, the connector extends down by 250 mm to make contact with a charging pad on the ground. It uses a magnetic guide and rotation of the connector unit to align the lowered interface with the contact points of the matrix pad. The connector unit is also covered by bellows to protect it from dirt and prevent any short-circuits. Voltage is applied only at the contact points that are covered by the bellows, and this selection is handled in the base unit. The power transmission and connector position are electronically monitored in the vehicle and on the pad to ensure a safe connection. The charging plate can be touched during the charging process, as only the contacts under the bellows are activated. That means the pad poses no danger to animals, which could crawl beneath the charging vehicle. If the lowered contact interface is moved or displaced, the charging process stops immediately to ensure overall safety. The comms between the pad and connector in the vehicle is via wi-fi to start and stop the charging once connected. In future this will use the comms protocols defined in the ISO 15118 standard developed for wireless charging. The pad on the ground needs a power and data connection, and connects to the grid using the OCPP standard protocol. One issue is keeping the pad clean to ensure a good connection with the vehicle’s contacts. Liquid and dirt are removed by a cleaning routine that is part of the rotation movement of the bellows, and snow and ice can be thawed away by running a small current though the contact pads. In the first-generation product, the system is designed to overcome a vehicle ground clearance of 250 mm; the limiting factor is the height of the vehicle. The ground clearance can be further increased to target special use-cases such as delivery vehicles with greater ground clearance, but these also have more space available for a longer contact unit. Other approaches make use of existing CCS charging connectors and cables, and are designed to work with any type of autonomous EV, from a passenger car to a delivery vehicle or even a large truck. To accommodate these different types of vehicles, with charging points in different places and at different heights, one prototype system uses a linear motor positioning system and computer vision. This requires the vehicle to park in a specific place, and the linear motors position the connector in the x and y directions, guided by the computer vision system that targets the open connector port. An arm with the CCS connector then extends out to the vehicle to make the connection. This can provide up to 300 kW of fast charging, say the developers, using the same charging and comms technologies implemented in modern fast chargers. The linear motors can handle any height or distance in the x and y directions and are limited only by the length of the cable connecting to the charger. For fast charging systems, these cables include cooling, so they’re heavy. For charging systems where people have to plug the connectors in, designers have had to reduce the length of the cables so that they are still useable, as the longer cables are too heavy to lift. This automated approach avoids this problem and so can support a wider range of vehicle designs. However, the challenge is in the computer vision system, which has to identify the charging port and its orientation in the dark or in rain, snow or fog. Sometimes the port can be tilted slightly upwards, and the vehicle can stop at an angle to the charging station, again changing the image. To solve that, the system uses a machine learning deep neural network (DNN). This DNN April/May 2022 | Unmanned Systems Technology This pinned connector system can be lowered from the underneath of a vehicle for automated charging (Courtesy of Easelink)