Simulation-Driven Evolutionary Motion Parameterization for Contact-Rich Granular Scooping with a Soft Conical Robotic Hand
Yongliang Wang, Cristian Camilo Beltran-Hernandez, Tomoya Takahashi, Masashi Hamaya
AI summary
Problem
Most robotic scooping tools are rigid and lack adaptability, while soft tools face complex control challenges and lack accurate simulation models, hindering efficient trajectory optimization and real-world deployment.
Approach
The authors build a MuJoCo simulation model of a deformable soft hand and use an evolutionary strategy to automatically optimize scooping trajectories and hand angles from visual input, which are then directly transferred to a physical robot.
Key results
- Physics-based MuJoCo simulation model of a soft deformable hand capturing passive reconfiguration dynamics
- Evolutionary trajectory optimization framework mapping visual container features to optimized scooping motions
- Successful sim-to-real transfer with minimal gap across diverse container sizes and granular materials
- Strong generalization to challenging tasks previously beyond rigid or fixed-shape tool capabilities
Why it matters
Enables adaptive, tool-free scooping for assistive feeding and material handling by bridging the gap between soft robotics, realistic simulation, and automated motion planning.
Abstract
Tool-based scooping is vital in robot-assisted tasks, enabling interaction with objects of varying sizes, shapes, and material states. Recent studies have shown that flexible, reconfigurable soft robotic end-effectors can adapt their shape to maintain consistent contact with container surfaces during scooping, improving efficiency compared to rigid tools. These soft tools can adjust to varying container sizes and materials without requiring complex sensing or control. However, the inherent compliance and complex deformation behavior of soft robotics introduce significant control complexity that limits practical applications. To address this challenge, this paper presents the development of a physics-based simulation model of a deformable soft conical robotic hand that captures its passive reconfiguration dynamics and enables systematic trajectory optimization for scooping tasks. We propose a novel physics- based simulation approach that accurately models the soft tool’s morphing behavior from flat sheets to adaptive conical structures, combined with an evolutionary strategy framework that automatically optimizes scooping trajectories without man- ual parameter tuning. We validate the optimized trajectories through both simulation and real-robot experiments. The re- sults demonstrate strong generalization and successfully address a range of challenging tasks previously beyond the reach of existing approaches. Videos of our experiments are available online: https://sites.google.com/view/scoopsh