钻井工程师依靠水力模型和钻屑输送模型来评估井眼清洁性能。然而，随着井深增加，这些模型的精度会下降，从而导致代价高昂的非生产时间 (NPT) 或卡钻事故。DrillDocs 公司的 CleanSight 是一款自动化振动筛监测系统，它利用嵌入边缘计算服务器和人工智能计算机视觉算法的工业摄像头，提取振动筛出口处钻屑的图像。该系统分析钻屑的大小和形状，然后报告可能表明井底钻具组合出现故障的异常物体。

3 月 17 日，在德克萨斯州加尔维斯顿举行的 IADC/SPE 国际钻井大会上，DrillDocs 的联合创始人兼首席执行官 Calvin Holt向DC介绍了 CleanSight 系统及其对实时决策的影响。

Drilling engineers rely on hydraulic models and drill cuttings transport models to measure hole-cleaning performance. However, these models can become more inaccurate as well depth increases, which can lead to costly nonproductive time (NPT) or stuck pipe incidents. CleanSight, an automated shale shaker monitoring system from DrillDocs, uses industrial cameras embedded with edge computing servers and AI-enabled computer vision algoritms to extract images of the cuttings as they exit the shale shaker. The system analyzes the size and shape of the cuttings, then reports anomalous objects that might indicate a part of the bottomhole assembly is failing.

Speaking to DC at the IADC/SPE International Drilling Conference in Galveston, Texas, on 17 March, Calvin Holt, Co-Founder and CEO of DrillDocs, talked about the CleanSight system and the impact it can have on real-time decision making.