Sensors & Payload

Kevin and Noodle use a common camera and localization architecture, while Kevin adds LiDAR and a dedicated capture mechanism for its RobotX competition tasks.

 

Camera

Kevin and Noodle both use an e-con e-cam86 camera designed for integration with the Jetson Orin Nano. Using a camera designed for the onboard computing platform simplifies integration and allows the team to focus on image processing and detection. The camera's lack of an external case also reduces its weight compared with standalone camera systems.

LIDAR

Kevin incorporates a RoboSense Airy LiDAR to provide additional environmental perception capabilities. The Airy was selected for its compact size and low weight, as well as its waterproof construction.

Kevin also reuses perception software developed for the USV, including GB-CACHE, to process and interpret LiDAR data. This allows the UAV to incorporate LiDAR information into its existing perception architecture.

GPS & IMU

Kevin and Noodle both use a Neo 3 GPS antenna along with the GPS receiver and IMU integrated into the Pixhawk flight controller. Using the Pixhawk's integrated localization hardware reduces vehicle weight and system complexity while relying on established components with extensive documentation.

The Neo 3 GPS antenna is designed for use with Pixhawk flight controllers, simplifying its integration with the UAV's flight-control system.

PAYLOAD SYSTEM

Capture Mechanism

Kevin incorporates a dedicated capture mechanism for retrieving and releasing the aluminum tins required for RobotX tasks.

The mechanism uses elastic bands to grip the tin through capstan force, providing a firm hold throughout the UAV's flight. A motor-driven mechanism uses a worm gear within the gearbox to prevent the mechanism from back driving when the motor is not powered.

The capture mechanism also allows some tolerance in the tin's position when the UAV lands to retrieve it, reducing the need for the tin to be perfectly centered beneath the vehicle.