Skip to content

frankkimrobotics/dynamo-grasp

 
 

Repository files navigation

DYNAMO-GRASP Project

This project introduces DYNAMO-GRASP, a novel approach to the challenge of suction grasp point detection. By leveraging the power of physics-based simulation and data-driven modeling, DYNAMO-GRASP is capable of accounting for object dynamics during the grasping process. This significantly enhances a robot's ability to handle unseen objects in real-world scenarios. Our method was benchmarked against established approaches through comprehensive evaluations in both simulated and real-world environments. It outperformed the alternatives by achieving a success rate of 98% in diverse simulated picking tasks and 80% in real-world, adversarially-designed scenarios. DYNAMO-GRASP demonstrates a strong ability to adapt to complex and unexpected object dynamics, offering robust generalization to real-world challenges. The results of this research pave the way for more reliable and resilient robotic manipulation in intricate real-world situations.

This repository provides the codebase for collecting data through simulation. These scripts will help users to generate their own datasets, further enhancing the extensibility and usefulness of DYNAMO-GRASP.

For more information please refer to our project website

For installation and usage instructions, please refer to the documentation

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • HTML 59.8%
  • Python 37.7%
  • C++ 1.9%
  • Other 0.6%