Suction-based Gripping and Sealing Data under Dynamic Robot Motions

This dataset collects sensor data in related to suction-based gripping during dynamic robot movements. The data are collected from an UR5e cobot arm and a suction-based gripping system based on Piab Picompact10x. The data are temporal sequences of dimensions include:

UR5e end-effector Tool Center Point (TCP) trajectories

UR5e end-effector wrench from F/T sensor

Commanded suction pressure from calibrated airflow regulator

Suction pressure measurement from picompact10x

The dataset evaluates how these readings change under various motion profiles, surface materials and inertia of the gripped objects. The motion patterns include straight, ciruclar, and star-shaped paths. Each pattern was executed under "low", "medium", "high" and "max" motion dynamics. Surface materials range from papercard, plastic film, sand papers to parameterized sinuoidal surfaces. 

Dataset format

Each trial is recorded as a numpy data file (.npy), with the file name indicating the parameters

{surface}_{motion_type}_{weight}g_{vel}v_{acc}a_{dynamics}.npy

An accompanied file with the suffix of "plot" visualizes the data sequence as an html file. 

Acknowledgement

This dataset collection is supported by the Innovation Fund Denmark (IFD) Grand Solutions project FERA - Fast and Efficient Robotic Automation via Re-Use of Data.