Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Towards Super-Nominal Payload Handling: Inverse Dynamics Analysis for Multi-Skill Robotic Manipulation

Anuj Pasricha, Alessandro Roncone

TL;DR

This work reframes payload limits as a design variable by showing that, for the Franka Emika Panda, dynamics-enriched planning can safely extend the operational envelope to around $2.7$ times the nominal payload. Using the Recursive Newton-Euler Algorithm to impose payload-aware torque constraints, and a large-scale data collection with cuRobo, the authors analyze both pick-and-place and pushing strategies. They find that non-prehensile primitives like pushing can substantially boost success rates at higher payloads, and that pushing and grasping are complementary in expanding the robot's manipulable space. The results motivate incorporating configuration-space awareness and mixed-motion strategies into planning frameworks to enable robust manipulation of super-nominal payloads in real-world tasks.

Abstract

Motion planning for articulated robots has traditionally been governed by algorithms that operate within manufacturer-defined payload limits. Our empirical analysis of the Franka Emika Panda robot demonstrates that this approach unnecessarily restricts the robot's dynamically-reachable task space. These results establish an expanded operational envelope for such robots, showing that they can handle payloads of more than twice their rated capacity. Additionally, our preliminary findings indicate that integrating non-prehensile motion primitives with grasping-based manipulation has the potential to further increase the success rates of manipulation tasks involving payloads exceeding nominal limits.

Towards Super-Nominal Payload Handling: Inverse Dynamics Analysis for Multi-Skill Robotic Manipulation

TL;DR

This work reframes payload limits as a design variable by showing that, for the Franka Emika Panda, dynamics-enriched planning can safely extend the operational envelope to around times the nominal payload. Using the Recursive Newton-Euler Algorithm to impose payload-aware torque constraints, and a large-scale data collection with cuRobo, the authors analyze both pick-and-place and pushing strategies. They find that non-prehensile primitives like pushing can substantially boost success rates at higher payloads, and that pushing and grasping are complementary in expanding the robot's manipulable space. The results motivate incorporating configuration-space awareness and mixed-motion strategies into planning frameworks to enable robust manipulation of super-nominal payloads in real-world tasks.

Abstract

Motion planning for articulated robots has traditionally been governed by algorithms that operate within manufacturer-defined payload limits. Our empirical analysis of the Franka Emika Panda robot demonstrates that this approach unnecessarily restricts the robot's dynamically-reachable task space. These results establish an expanded operational envelope for such robots, showing that they can handle payloads of more than twice their rated capacity. Additionally, our preliminary findings indicate that integrating non-prehensile motion primitives with grasping-based manipulation has the potential to further increase the success rates of manipulation tasks involving payloads exceeding nominal limits.
Paper Structure (6 sections, 2 figures)

This paper contains 6 sections, 2 figures.

Table of Contents

  1. Motivation
  2. Payload-Aware Torque Constraints
  3. Data Collection
  4. Pick-and-Place Trajectory Analysis
  5. Push Trajectory Analysis
  6. Conclusion

Figures (2)

  • Figure 1: Despite a decrease in dynamically-valid trajectories for increasing payloads, the robot maintains significant functionality at 2.7 times the nominal payload of 3 $kg$. It retains a reachable area of 0.99 $m^2$ at 8 $kg$, only a slight decrease from 1.06 $m^2$ at 3 $kg$.
  • Figure 2: When pushing, the robot maintains high percentages of dynamically-valid trajectories up to 15 $kg$, with a significant drop beyond (top left). Even at 2.7 times the nominal payload of 3 $kg$, the robot retains substantial functionality, maintaining a reachable area of 0.99 $m^2$ at 8 $kg$, only slightly reduced from 1.00 $m^2$ at 0 $kg$.