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

55 Its sensor architecture for autonomous traffic detection and navigation uses four EO cameras, three Lidars, three radars and a GNSS receiver. The company notes that extensive redundancy was needed in the sensors owing to the size of the bus and the shadow it can cast, making it challenging to achieve a full surrounding view at all times. Two cameras face the front, with two others facing the rear from the sides. The three Lidars also face the front (with the outer two covering close to a 270° field of view), while one of the three radars faces the front and two face the rear from the side of the bus This ensures point clouds are created with sufficient speed and accuracy to gauge traffic dynamics and ranges directly in front of the bus, while the cameras’ computer vision systems identify and classify objects surrounding the bus, in order to predict their likely movements. The radars account for vehicles approaching from the rear. This is an important consideration, given the possibility of collisions from the rear, but not life-threatening to humans in the way that forward collisions could be, so power-hungry Lidars are not used here. “The GNSS meanwhile is the primary source of mapping and localisation data,” Rego adds. “And if there’s sunlight or fog, these can obscure the cameras’ work, which is why Lidar and radar are our primary tools for object detection and rangefinding.” The autonomous mode is engaged via an electromagnet that clutches the Autonomous transport | Insight of the matter An ie bus from Irizar E-Mobility has begun commercial trials in Malaga, Spain, to SAE Level 3 autonomy (Courtesy of Irizar Group) Unmanned Systems Technology | December/January 2022 Rory Jackson looks at some of the companies who are turning the hype surrounding autonomous transport into practical systems

RkJQdWJsaXNoZXIy MjI2Mzk4