Technology Focus: Data Science, Analytics, and Artificial Intelligence (January 2025)
Bibliographic record
Abstract
Almost every day, petroleum engineers are coming to realize that they’ve got an arsenal of good ideas on how to leverage large, messy data sets to add value to their businesses. Those who have enlisted in the Analytics Army have progressed from siloed digitalization attempts to well-concerted digital transformation strategies that reflect high levels of organizational digital maturity. Paper SPE 220686 hits an undeniable sweet spot for production and reservoir engineers, using an array of machine-learning methods to accurately predict real-time well status. Practicality here is evident, because only widely available surface-measured pressures, temperatures, and choke-valve positions are used to classify online and offline well status. The 99% accuracy levels achieved represent a boon not only for those interested in more-reliable rate allocation but also for well-integrity and flow-assurance applications. With ever-increasing data acquisition volumes, data-labeling and categorization problems increase in lockstep. After all, how well can your luggage be found without proper tagging? Anyone remember the 2022 summer of lost luggage? The value proposition of paper SPE 218865 is clear, using a combination of natural language processing for coiled tubing operations reports and pattern recognition of multimodal job data to automatically label the job types and technologies used. A myriad of other cases spring to mind for repurposing the solution to transform current onerous processes of metadata generation. In the aftermath of the explosion in popularity of all things generative artificial intelligence, paper SPE 217671 steals the limelight. This paper exploits good data foundations by painting a roadmap to building your own drilling chatbot adviser. Compared with other generalized large language models, the zero-shot learning technique allows the chatbot to answer queries for drilling domain-specific knowledge that it hasn’t explicitly “seen.” The authors have even publicized the training data sets collated from the Norwegian Petroleum Directorate to enable quick and relatively inexpensive replication. Recommended additional reading at OnePetro: www.onepetro.org. URTeC 4045912 Using Machine Learning To Automate Fracture-Driven-Interaction Analysis by Reid Thompson, Momentum AI, et al. SPE 220833 Preliminary Research on Applications of Large Language Models in the Exploration and Production Industry by X.G. Zhou, PetroChina, et al. SPE 220714 Automated Well and Reservoir Management Using Hybrid Physics and Data-Driven Models—Case Study by Azreen Mustafa, Hess Corporation, et al.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.009 | 0.015 |
| Meta-epidemiology (narrow) | 0.003 | 0.001 |
| Meta-epidemiology (broad) | 0.002 | 0.001 |
| Bibliometrics | 0.006 | 0.006 |
| Science and technology studies | 0.002 | 0.004 |
| Scholarly communication | 0.017 | 0.020 |
| Open science | 0.003 | 0.007 |
| Research integrity | 0.011 | 0.010 |
| Insufficient payload (model declined to judge) | 0.214 | 0.222 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".