xpertsystems/oil012-sample

OIL-012是一个合成的钻井平台传感器物联网数据集（样本），专为石油和天然气上游领域的预测性维护机器学习、状态监测机器学习、传感器融合建模和剩余使用寿命预测而设计。该样本涵盖了来自10个全球盆地和8种钻井平台类型的30个钻井平台，包含134,622行数据，其中94,745行为稀疏格式的遥测事件，分布在15个表格中。数据包括钻井参数、振动分析、液压系统、电力系统、热监测、传感器健康状态、报警事件、维护记录、剩余使用寿命标签等，并基于API、ISO、IEEE等行业标准进行校准，确保物理一致性和行业真实性。数据集适用于多种机器学习任务，如回归、分类、异常检测和关系型机器学习。OIL-012 is a synthetic rig sensor IoT dataset (sample) designed for predictive-maintenance ML, condition-monitoring ML, sensor-fusion modeling, and remaining-useful-life forecasting in the oil and gas upstream sector. The sample covers 30 rigs across 10 global basins and 8 rig types, with 134,622 rows including 94,745 sparse-format telemetry events linked across 15 tables. Data includes drilling parameters, vibration analysis, hydraulic systems, power systems, thermal monitoring, sensor health, alarm events, maintenance records, remaining useful life labels, and more, calibrated to industry standards such as API, ISO, and IEEE for physical consistency and industry authenticity. The dataset is suitable for various ML tasks including regression, classification, anomaly detection, and relational machine learning.