Scholarlink

<p>This paper introduces a novel framework for optimal dexterous manipulation of three-dimensional (3D) micro-objects with complex shapes through four robotized probes. The framework utilises a geodesic polyhedron sphere sampling method within the objects’ configuration space to construct original graphs of interconnected stable, accessible, and collision-free grasps. The impact of this sampling method on the generated probe paths was evaluated, with performance assessed in terms of cost and time efficiency. To ensure robust manipulation paths, a worst-case scenario approach was adopted, where the dominant and time-varying adhesion forces at the micro-scale were overestimated when acting as a disturbing factor and disregarded when contributing to stabilisation. The A* algorithm was employed to generate complete and optimal manipulation probe paths. The simulation results demonstrated, for the first time, collision-free manipulation planning of miniaturised 3D objects with complex shapes in the presence of adhesion forces.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Dexterous manipulation planning of 3D miniaturized objects by four robotized probes in the presence of adhesion forces

  • 期刊论文

  • 发表日期：

    2025年9月5日

查看全文
  • Ivan Jordan Tchouatat Kepseu,
  • Tala Dannawi Aissaoui,
  • Michael Gauthier,
  • Redwan Dahmouche

摘要

This paper introduces a novel framework for optimal dexterous manipulation of three-dimensional (3D) micro-objects with complex shapes through four robotized probes. The framework utilises a geodesic polyhedron sphere sampling method within the objects’ configuration space to construct original graphs of interconnected stable, accessible, and collision-free grasps. The impact of this sampling method on the generated probe paths was evaluated, with performance assessed in terms of cost and time efficiency. To ensure robust manipulation paths, a worst-case scenario approach was adopted, where the dominant and time-varying adhesion forces at the micro-scale were overestimated when acting as a disturbing factor and disregarded when contributing to stabilisation. The A* algorithm was employed to generate complete and optimal manipulation probe paths. The simulation results demonstrated, for the first time, collision-free manipulation planning of miniaturised 3D objects with complex shapes in the presence of adhesion forces.