Robotic Arm Dataset (RoAD)

该数据集名为RoAD，包含了在实际生产过程环境中，安装在Kuka LBR iiwa机器人臂上的多变量时间序列传感器数据。这些数据来自加速度计、陀螺仪和功耗传感器，并附有“异常”标签，用以指示是否存在碰撞、重量和速度异常等不同类型的异常情况。此外，数据集还包含了一份元数据描述，将机器人操作手执行的每个动作以0到30的数值进行分类。该数据集旨在支持机器学习驱动算法的开发与测试，其规模源自配备有协作型机器人臂的完整生产线，适用于执行网络物理系统（CPSs）中的故障诊断和异常检测任务。

This dataset is named RoAD. It contains multivariate time-series sensor data collected from a Kuka LBR iiwa robotic arm installed in an actual production environment. The sensor data originates from accelerometers, gyroscopes, and power consumption sensors, and is annotated with "anomaly" labels that indicate different types of abnormal conditions such as collisions, abnormal weight and abnormal speed. In addition, the dataset includes a metadata description that categorizes each action performed by the robotic manipulator with numerical values ranging from 0 to 30. This dataset aims to support the development and testing of machine learning-driven algorithms. Derived from a complete production line equipped with collaborative robotic arms, it is applicable for fault diagnosis and anomaly detection tasks in cyber-physical systems (CPSs).

搜集汇总

背景与挑战

背景概述

Robotic Arm Dataset (RoAD)是一个用于基于过程挖掘的机械臂传感器数据建模和仿真的数据集，支持异常检测、事件日志提取和Petri网仿真等实验。数据集需要特定Python环境和库支持。