Automatic Daily Drilling Mud Report Processing Using Generative AI to Maximize the Operational Efficiency

Abstract Large service companies process an excessive amount of drilling mud reports daily, requiring engineers to perform labor-intensive, costly, and error-prone manual analysis work. Generative AI offers an ideal solution to automate this routine task. This study proposes an innovative yet resource-efficient mud report processing framework using generative AI. Within this framework, an automated pipeline is established to capture and process daily mud reports from varied sources such as emails or pdfs. Mud reports are tabular-rich document and its contents are extracted using Optical Character Recognition (OCR) technologies and Generative AI. The extracted data is stored in structured databases, and then visualized on an interactive business intelligence (BI) dashboard to generate business values and insights. Observations confirm that the proposed method efficiently handles daily drilling mud reports while maintaining near-perfect accuracy with negligible computational time and high consistency of the results. The architecture of the model is designed to effectively handle reports from various existing drilling fluid vendors. Furthermore, it is built to process reports from new, previously unseen vendors in a plug-and-play manner, without requiring any modifications to the existing model. The system offers full transparency in measuring operational efficiency and cost, processing hundreds of mud reports in a fraction of the time compared to traditional manual methods. Detailed analysis shows that the implementation of Generative AI has improved processing efficiency by reducing the time required per report by 99%, resulting in a significant boost in overall productivity. In conclusion, this Generative AI architecture offers a reliable, cost-effective, and scalable solution, revolutionizing the automation of large-scale daily drilling mud report processing within the energy industry. Additionally, methods for estimating Generative AI costs and operational cost reductions are discussed, further highlighting the potential for profitability in the petroleum industry through Generative AI-driven automation.