# AURSAD: Universal Robot Screwdriving Anomaly Detection Dataset

arXiv: https://arxiv.org/abs/2102.01409 · HTML: https://mlsanity.com/paper/2102.01409/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Builds on 1 of its 5 references that are in the map.
- First posted on arXiv 2 Feb 2021.

## Builds on (most cited first, 1 of 1)

- [Temporal Attention augmented Bilinear Network for Financial Time-Series Data Analysis](https://mlsanity.com/paper/1712.00975/) (2017)

## Built on by (most cited first, 5 of 5)

- [FactoryNet: A Large-Scale Dataset toward Industrial Time-Series Foundation Models](https://mlsanity.com/paper/2605.09081/) (2026)
- [Detecting Faults during Automatic Screwdriving: A Dataset and Use Case of Anomaly Detection for Automatic Screwdriving](https://mlsanity.com/paper/2107.01955/) (2021)
- [FactoryBench: Evaluating Industrial Machine Understanding](https://mlsanity.com/paper/2605.07675/) (2026)
- [DeepTimeAnomalyViz: A Tool for Visualizing and Post-processing Deep Learning Anomaly Detection Results for Industrial Time-Series](https://mlsanity.com/paper/2109.10082/) (2021)
- [PyScrew: A Comprehensive Dataset Collection from Industrial Screw Driving Experiments](https://mlsanity.com/paper/2505.11925/) (2025)

## Abstract (arXiv)

Screwdriving is one of the most popular industrial processes. As such, it is increasingly common to automate that procedure by using various robots. Even though the automation increases the efficiency of the screwdriving process, if the process is not monitored correctly, faults may occur during operation, which can impact the effectiveness and quality of assembly. Machine Learning (ML) has the potential to detect those undesirable events and limit their impact. In order to do so, first a dataset that fully describes the operation of an industrial robot performing automated screwdriving must be available. This report describes a dataset created using a UR3e series robot and OnRobot Screwdriver. We create different scenarios and introduce 4 types of anomalies to the process while all available robot and screwdriver sensors are continuously recorded. The resulting data contains 2042 samples of normal and anomalous robot operation. Brief ML benchmarks using this data are also provided, showcasing the data's suitability and potential for further analysis and experimentation.