Detecting Faults during Automatic Screwdriving: A Dataset and Use Case of Anomaly Detection for Automatic Screwdriving - 2021-07-05 - Research Publication - The study trains TABL and ResNet time-series classifiers on 125 robot and screwdriver features from 2,045 labeled normal and faulty OnRobot Screwdriver operations. Across four-way fault classification, ResNet achieved an… - source: arxiv.org
AURSAD: Universal Robot Screwdriving Anomaly Detection Dataset - 2021-02-02 - Research Publication - Researchers paired a UR3e robot with the OnRobot Screwdriver and continuously recorded all available robot and screwdriver sensors through normal operation and four deliberately introduced fault scenarios. The resulting… - source: arxiv.org
AURSAD: Universal Robot Screwdriving Anomaly Detection Dataset - 2021-02-01 - Research Publication - The open Zenodo release contains 2,045 labeled screwdriving samples recorded at 100 hertz from a UR3e robot and OnRobot Screwdriver, distributed as 1,420 normal and 625 faulty operations. Fault labels cover damaged screws… - source: zenodo.org