TreeSpider

TreeSpider: In-canopy exploration with tether-based aerial modular arms

Aerial robots offer significant potential for forest canopy research and environmental monitoring by enabling data collection at high spatial and temporal resolutions. However, their limited maneuverability constrains their ability to navigate and access dense canopy environments. To address this challenge, we propose a tether-based drone that integrates perching arms with anchoring capability for in-canopy stabilization and manipulation. Once tethered to a branch, our system, equipped with a 360° ring, can maneuver freely within the canopy in any direction.

The ring mechanism decouples the tether from the drone frame, allowing the pitch angle to be reoriented independently along the longitudinal axis. Additionally, the system adjusts the tether length dynamically to optimize movement and conserve energy. This flexibility allows the drone not only to approach branches from various angles within dense foliage but also to extend its reach to adjacent trees while remaining tethered. By hanging from one tree to approach and sense another, safe and efficient multi-tree sensing is facilitated.

The perching arms can then also wrap around tree branches for secondary perching, allowing for physical interaction with the environment at multiple locations over long time scales. Our experiments demonstrated that tethered operation can reduce energy consumption by up to fivefold compared to untethered hovering. The proposed system incorporates a leaf sampling tool and a winding mechanism to adjust tether length. Additionally, testing of the perching arms showed reliability, with retention forces increasing as the winding angle increased, achieving up to 10 N of retention force, thereby enhancing resistance to falling from branches.

This system advances the state-of-the-art in forest robotic sensing and provides an efficient solution for maneuvering within the canopy.

[1] TreeSpider: In-canopy exploration with tether-based aerial modular arms, Submitted to IEEE Robotics and Automation Letters (RA-L), Romanello L., Lan T., S. Hassan, M. Kovac, S. F. Armanini, B. B. Kocer