Authors:

Jihyo Park

Date Created:

Not specified

Course Title:

Professor:

Not specified

About Paper:

Agricultural drain tiles, spanning a network of 56 million hectares in the Midwest and Southern United States, are essential for agricultural productivity. These tiles, buried six feet below the surface, offer a unique gateway to the subterranean environment. Despite their extensive reach and significant impact, effective tools for directly mapping and monitoring drain tiles are lacking. The corrugated structure and uneven terrain of these tiles necessitate adaptive locomotion capabilities. Traditional vision- and LiDAR-based systems for mapping and locomotion are limited by the feature-sparse and repetitive nature of the drain tile environment. In this research, we present a novel robotic platform specifically designed for the drain tile environment, utilizing wheg (wheel with legs) locomotion. We employ a central pattern generator with Hopf oscillators to create rhythmic movements of the whegs, enabling them to mesh with the corrugations for enhanced traction. This neuromorphic architecture is deployed directly on the robot, providing superior autonomous control in the challenging drain tile environment. We evaluate the effectiveness of our approach against standard robotics SLAM methods, including ORB-SLAM3 and monocular SLAM, as implemented in standard ROS packages. Performance metrics such as system footprint, latency, and overheads are compared. This work demonstrates the practical feasibility and effectiveness of our subterranean robot for large-scale monitoring of drain tiles. Overall, this research introduces a novel robot and system for studying the drain tile environment, illuminating the subterranean soil microbiome in unprecedented ways. Keywords: [no keywords provided]

Source:

Purdue University / 2024

Topics:

No topics listed

Co-authors:

Jihyo Park