Contact-Rich Motion Planning via GPU-Parallel Mode Evaluation
研究通过GPU并行模式评估解决接触丰富运动规划的计算难题，提出CoMET方法，在平面推力基准测试中表现出色。
0 citationsRead paper
Selected work
Representative Papers
FOCAL-VLA: Subtask-Guided Geometry Distillation and Implicit World Modeling for Vision-Language-Action Models
为解决VLA模型在精确和长期操作中的局限性，提出FOCAL-VLA框架，通过子任务引导的几何蒸馏与隐式世界建模来提高模型的空间及时序理解能力。
0 citationsRead paper
ForwardDLO: Model-Based Bimanual Shape Matching of Unconstrained Deformable Linear Objects
研究提出ForwardDLO模型，通过预测绳索各段位移解决机器人双臂操控非固定柔性线性物体形状匹配问题。
0 citationsRead paper
Sequential Object Placement Optimization with Convex Decomposition
本文通过引入SOPO-CD框架，将物体放置问题转化为可微非线性优化问题，解决了机器人在处理不规则形状物体时的高效打包难题。
0 citationsRead paper
Coordinated Multi-Robot Disassembly for Makespan Optimization of Large-Scale Assemblies
This work addresses the challenge of prolonged makespan in multi-robot collaborative disassembly within confined spaces, where motion conflicts frequently occur. The authors propose CoMuDi, a novel approach that deeply integrates spatiotemporal RRT* (ST-RRT*) into multi-robot disassembly planning for the first time. CoMuDi models the assembly using a dependency graph to generate composite tasks, propagates temporal constraints to coordinate robot actions, and leverages ST-RRT* to optimize the execution time of individual tasks, thereby minimizing overall makespan. Evaluated across six benchmark scenarios involving up to 49 parts and nine robots, CoMuDi significantly improves planning success rates while effectively reducing both makespan and robot idle time.
0 citationsRead paper
Recent publications
Latest Papers
Contact-Rich Motion Planning via GPU-Parallel Mode Evaluation
研究通过GPU并行模式评估解决接触丰富运动规划的计算难题，提出CoMET方法，在平面推力基准测试中表现出色。
0 citationsRead paper
FOCAL-VLA: Subtask-Guided Geometry Distillation and Implicit World Modeling for Vision-Language-Action Models
为解决VLA模型在精确和长期操作中的局限性，提出FOCAL-VLA框架，通过子任务引导的几何蒸馏与隐式世界建模来提高模型的空间及时序理解能力。
0 citationsRead paper
ForwardDLO: Model-Based Bimanual Shape Matching of Unconstrained Deformable Linear Objects
研究提出ForwardDLO模型，通过预测绳索各段位移解决机器人双臂操控非固定柔性线性物体形状匹配问题。
0 citationsRead paper
Sequential Object Placement Optimization with Convex Decomposition
本文通过引入SOPO-CD框架，将物体放置问题转化为可微非线性优化问题，解决了机器人在处理不规则形状物体时的高效打包难题。
0 citationsRead paper
Coordinated Multi-Robot Disassembly for Makespan Optimization of Large-Scale Assemblies
This work addresses the challenge of prolonged makespan in multi-robot collaborative disassembly within confined spaces, where motion conflicts frequently occur. The authors propose CoMuDi, a novel approach that deeply integrates spatiotemporal RRT* (ST-RRT*) into multi-robot disassembly planning for the first time. CoMuDi models the assembly using a dependency graph to generate composite tasks, propagates temporal constraints to coordinate robot actions, and leverages ST-RRT* to optimize the execution time of individual tasks, thereby minimizing overall makespan. Evaluated across six benchmark scenarios involving up to 49 parts and nine robots, CoMuDi significantly improves planning success rates while effectively reducing both makespan and robot idle time.
0 citationsRead paper