Autonomous Multi-legged Robot for Crop/Turf Management

Agriculture is resource intensive; monitoring can make it more efficient. However, traditional monitoring techniques are either disruptive or very inefficient. This project seeks to address these challenges by using an autonomous spider robot to use RGB-IR (RGB and infrared) camera and GPS data with existing plant health proxies to determine the health of a field.

Polyethylene terephthalate glycol 3D-printed plastic pieces and servo drivers allow the eight-legged robot to use a custom spider-inspired locomotion to walk 0.15 km per hour and rotate 360 degrees as needed. The robot is controlled by a microcomputer that uses a machine learning camera, GPS, and a nine degrees of freedom sensor to follow waypoints the use inputs through Google My Maps. The microcomputer instructs the robot to walk along the path set by the waypoints, avoiding obstacles the camera detects. The robot can operate in extreme heat conditions and in light rain for about one hour. As it walks, the robot passively collects RGB-IR data that it sends to a repository where it is processed. The user can then use the data to best manage the field.