TADI: AI system optimises drilling data for the oil industry
A new AI system named TADI has been developed to analyse complex drilling data within the oil and gas industry. The system aims to improve decision-making by integrating and processing heterogeneous data sources.
Vad har hänt
Researchers have presented TADI (Tool-Augmented Drilling Intelligence), an agent-based AI system designed to transform operational drilling data into evidence-based analytical information. The system has been applied to Equinor's Volve Field data and integrates 1,759 daily drilling reports, selected real-time WITSML objects, 15,634 production records, formation data, and perforations. TADI employs a dual architecture using DuckDB for structured queries across 12 tables with 65,447 rows, and ChromaDB for semantic search across 36,709 embedded documents.
Key facts
| Antal dagliga borrningsrapporter | 1 759 |
|---|---|
| Antal produktionsposter | 15 634 |
| Antal rader i DuckDB | 65 447 |
| Antal inbäddade dokument i ChromaDB | 36 709 |
| Antal verktyg orkestrerade av LLM | 12 |
| Antal automatiska tester | 95 |
”We present TADI (Tool-Augmented Drilling Intelligence), an agentic AI system that transforms drilling operational data into evidence-based analytical intelligence.”
”Applied to the Equinor Volve Field dataset, TADI integrates 1,759 daily drilling reports, selected WITSML real-time objects, 15,634 production records, formation tops, and perforations into a dual-store architecture.”
”The system parses all 1,759 DDR XML files with zero errors, handles three incompatible well naming conventions, and is backed by 95 automated tests plus a 130-question stress-question taxonomy spanning six operation”
Varför det spelar roll
TADI addresses the challenge of heterogeneous data sources and incompatible naming conventions, which have historically hindered effective data analysis in the oil and gas sector. By orchestrating twelve domain-specialised tools via a large language model (LLM), the system can cross-reference structured drilling measurements with narratives from daily reports. This enables more comprehensive and fact-based intelligence for operational decisions.
Vem påverkas
The system primarily affects stakeholders in the oil and gas industry, particularly those managing complex drilling operations and large datasets. Researchers and AI developers can also benefit from TADI's design to develop similar agent-based systems for other domains. Equinor is specifically mentioned as the data source used for validation.
EU-status
TADI has been applied to data from Equinor's Volve Field, a Norwegian oil field. The system's potential to streamline drilling operations is relevant to the energy sector within the EU, though implementation depends on local regulations and data standards. While Norway is not a member of the EU, it maintains close ties to the EU energy market via the EEA Agreement.
Mer att veta
The system has successfully interpreted all 1,759 DDR XML files without errors and handles three incompatible well-naming conventions. It is supported by 95 automated tests and a taxonomy of 130 stress questions across six operation types, reinforcing its robustness and reliability.
Quick answers about this story
Vad har hänt?
När hände det?
Varför spelar det roll?
Vilka bolag berörs?
Länken öppnar i nytt fönster och leder till utgivarens egen sida.
Källan har spårats automatiskt från utgivaren via Aheadlines signalkedja.