IKLink

Implementation of our ICRA'24 paper: IKLink: End-Effector Trajectory Tracking with Minimal Reconfigurations

[Pre-print] [Supplementary Video]

Introduction

IKLink enables a robot manipulator to track reference end-effector trajectories of any complexity while performing minimal reconfigurations. IKLink eliminates the need to manually segment a long or complex trajectory and is beneficial in real-life scenarios that involve end-effector trajectory tracking, such as welding, sweeping, scanning, painting, and inspection. For more information, please refer to our paper.

Getting Started

1. Install Rust

2. Compile:

   cargo build

3. Run a demo:

   cargo run --bin traj_tracing

   The demo processes end-effector trajectories in input_trajectories and saves the generated motions in output_motions.

   We note that this Rust implementation is about 10x fasters than the Python implementation described in the paper.

4. Expected output:

   Constructing nodes for point 0 / 767
   Constructing nodes for point 1 / 767
   ...
   Constructing nodes for point 766 / 767
   Running dynamic programming algorithm
   Min Num of Reconfig: 2
   Saved motion to: <some_dir>/iklink/output_motions/panda_2023-08-25_11-37-55.csv
   

Supplementary Video

YouTube video link

Citation

@inproceedings{wang2024iklink,
  title={ IKLink: End-Effector Trajectory Tracking with Minimal Reconfigurations},
  author={Wang, Yeping and Sifferman, Carter and Gleicher Michael},
  booktitle={2024 IEEE International Conference on Robotics and Automation (ICRA)},
  year={2024},
  organization={IEEE}
}