Jointly satisfying reachability and collision-avoidance constraints

Develop a solution that more effectively addresses both hand reachability and arm–environment collision avoidance during dexterous grasp generation near robotic-arm workspace boundaries, while improving the associated constraint-satisfaction metrics.

Background

The paper evaluates grasp generation for objects placed near the edge of the robotic arm’s workspace, where only certain wrist orientations are reachable and collision avoidance with the tabletop is also required. Ablation experiments that apply only collision avoidance or only hand reachability optimize their respective metrics but substantially impair the other constraint. The authors’ combined guidance method improves the overall feasible-grasp rate, yet explicitly acknowledge that a more effective solution balancing both constraints has not been found.

References

However, this comes at the expense of the other metrics, indicating that a solution addressing both constraints more effectively remains to be found.

Arm-Aware Guided Dexterous Grasp Generation with Arm-Agnostic Grasp Models  (2608.16351 - Jia et al., 17 Aug 2026) in Section 4.1.2, subsection “Performance of the Hand Reachability Constraint”