New Role of Technical Specialists to Enable Digital Transformation in the Petroleum Industry: A Petrophysicist-Based Proof of Concept
The DOI matches its Crossref record; 6 of the 6 references it cites were found in a registry or an index.
Abstract
Standard petrophysical workflows are limited by centralized control of log data, restricting tool access to dedicated specialists. This structure often leads to delays, reduced reproducibility and limited accessibility for multidisciplinary teams. To address these challenges, this study proposes a methodological framework combining natural language interfaces with an GenAI-based orchestration layer. The approach allows non-petrophysical specialists to access key petrophysical data, perform standard analyses, reduce reliance on manual processing and ensure secure, reproducible operations. The system is built on a modular architecture to streamline and standardize petrophysical data analysis. The Natural Language Interface uses large language models to interpret user queries and translate them into structured tool commands. The MCP Server Execution layer orchestrates analytical workflows, manages access control, and coordinates the execution of core analytical tools. Library Operations provide the fundamental well log processing, quality control, statistical analysis, and data visualization. The containerized system supports local deployment, ensuring traceability and data confidentiality within secure corporate infrastructure. Results indicate that the proposed agent-based framework mitigates procedural constraints associated with centralized log interpretation, facilitating more autonomous access for subsurface teams. The natural language interface translates user queries into structured analytical sequences, supporting reproducible execution and preserving auditability. This outcome supports statistical summaries, anomaly indicators, and customized visualizations optimized for interpretation by multidisciplinary teams. The framework enables petrophysicists to transition from task execution to system-level oversight, focusing on the design, validation, and supervision of automated workflows. Data security is a critical concern, addressed through containerized local deployment and the use of private LLM instances, ensuring compliance with internal data governance policies and eliminating external risks. The system integrates internal corporate data with locally hosted open-source libraries deployed under strict network isolation to ensure that no data is transmitted outside the organizational environment. The results confirm the feasibility of the approach as a scalable solution for restructuring interpretation workflows and reducing barriers in geoscientific collaboration. This study introduces a secure framework for embedding natural language-driven AI agents into subsurface workflows. A key impact is the shift in the petrophysicist's role from repetitive execution to workflow supervision and agent training. Significant business impact by reducing workload and decisions acceleration while supporting full data control and business process modularity. The cognitive layer uses structured prompting, including chain-of-thought reasoning to enhance interpretability and robustness of responses.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Citations
Cites
Chain-Of-Thought Prompting Elicits Reasoning in Large Language Models
Wei, J., Wang, X., Schuurmans, D., Bosma, M., Ichter, B., Xia, F., Chi, E., Le, Q. V., & Zhou, D. (2022) · Advances in Neural Information Processing Systems 35 · doi:10.52202/068431-1800 · Cited by 4,505
Liu, P., Yuan, W., Fu, J., Jiang, Z., Hayashi, H., & Neubig, G. (2023) · ACM Computing Surveys · doi:10.1145/3560815 · Cited by 4,361
Unleashing the potential of prompt engineering for large language models
Chen, B., Zhang, Z., Langrené, N., & Zhu, S. (2025) · Patterns · doi:10.1016/j.patter.2025.101260 · Cited by 490
Applying Machine Learning NLP Algorithm for Reconciliation Geology and Petrophysics in Rock Typing
Tveritnev, A., Khanji, M. Y., Abdullah, S., Rojas, L., Ermilov, A., Mansoori, F. H. A., & Alblooshi, A. (2023) · Abu Dhabi International Petroleum Exhibition & Conference (ADIPEC) · doi:10.2118/216223-ms · Cited by 4
Self-citationGuest Editorial: 2025 Perspective: Digital Transformation in Oil and Gas—From Evolution to Shockwave
Thambynayagam, M. (2025) · Journal of Petroleum Technology · doi:10.2118/0325-0007-jpt · Cited by 2
Bi, Python and Low Code Applications to Accelerate Digital Transformation
Heine, S. J., Castillo, O. A., Palacios, M. R., López, C., Castello, M. E., Albarracin, H. R., & Serrano, G. (2023) · SPE Latin American and Caribbean Petroleum Engineering Conference · doi:10.2118/213108-ms · Cited by 1
Research works printed but not linked: 6
- Vaswani, 2017, Attention is All you Need
- Inverarity, 2025, Lasio: Python library for reading and writing well data using Log ASCII Standard (LAS) files
- Hunter, 2007, Matplotlib: A 2D graphics environment, 90
- Wang, X., Wei, J., Schuurmans, D., Le, Q., Chi, E., Narang, S., … Zhou, D. (2023, Mar7). [2203.11171] Self-Consistency Improves Chain of Thought Reasoning in Language Models. Retrieved from arXiv.org e-Print archive: https://arxiv.org/pdf/2203.11171 · arXiv
- Brown, T., Mann, B., Ryder, N., Subbiah, M., Kaplan, J., Dhariwal, P., … Shyam, P. (2020). Language Models are Few-Shot Learners. Retrieved from NeurIPS Proceedings: https://papers.nips.cc/paper/2020/hash/1457c0d6bfcb4967418bfb8ac142f64a-Abstract.html
- Jurafsky, D., & Martin, J. H. (2008). ResearchGate. Retrieved from (PDF) Speech and Language Processing: An Introduction to Natural Language Processing, Computational Linguistics, and Speech Recognition: https://www.researchgate.net/publication/200111340_Speech_and_Language_Processing_An_Introduction_to_Natural_Language_Processing_Computational_Linguistics_and_Speech_Recognition
Cited by
Data Quality in the Age of AI: A Review of Governance, Ethics, and the FAIR Principles
Guillen-Aguinaga, M., Aguinaga-Ontoso, E., Guillen-Aguinaga, L., Guillen-Grima, F., & Aguinaga-Ontoso, I. (2025) · Data · doi:10.3390/data10120201 · Cited by 37
“Traditional frameworks, developed in an era when data primarily served human analysts [222], must now accommodate systems in which algorithms themselves are the primary consumers of data [223–225].”
Metadata
- Authors
- Ermilov, A., Tveritnev, A., & Trusova, A.
- Affiliations
- UiT The Arctic University of Norway (NO)
- Abu Dhabi National Oil (United Arab Emirates) (AE)
- Industrial University of Tyumen (RU)
- ORCID
- Tveritnev, A.
- Venue
- ADIPEC
- Publisher
- SPE
- Published
- 2025-11-03
- OpenAlex
- W4415777366
- Cited by
- 1
- References
- 5
- Citations by year
- 2025: 1
- FWCI
- 3.03
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- Abu Dhabi International Petroleum Exhibition & Conference (ADIPEC) (publishedVersion)
- Topics
- Social Sciences › Decision Sciences › Information Systems and Management › Scientific Computing and Data Management
- Physical Sciences › Computer Science › Artificial Intelligence › Artificial Intelligence Applications
- Physical Sciences › Earth and Planetary Sciences › Geology › Hydrocarbon exploration and reservoir analysis
- Keywords
- petrophysical interpretation, natural language interfaces, large language models, generative AI, workflow orchestration, well log interpretation, digital transformation, oil and gas industry, containerized deployment, data security, reproducibility, multidisciplinary collaboration, chain-of-thought prompting
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.