Target Handling Modalities with Obstacle Avoidance for Planar Soft Growing Manipulator Design

Abstract

Soft growing robots mimic plant-like growth to navigate complex environments thanks to their specific actuation and material. This class of robots can also be used for manipulation tasks. While manufacturing these robots for specific tasks, it is crucial to carefully design their length and placement of joints. In this work, we extend our state-of-the-art optimizer for planar soft growing manipulators design, which retrieves the optimal robot dimensions for a specific given task. While the first version of the optimizer only considered a base case (where targets were only points in space), in this work, we implement five target handling modalities based on real-case manipulation scenarios. Specifically, targets are treated as obstacles and, as such, occupy space in the environment. Depending on the modality, the way these targets are handled can change. Results show that with this extension, the optimizer can tackle different manipulation cases correctly.