Skip to main content
Default image description

DDBot – AI for automatic population of semantics

A tool that leverages Large Language Models (LLMs) to automatically populate semantic structures and streamline the configuration of D-WIS systems.

Challenge

D-WIS has been demonstrated for interoperability between automated systems of different companies. D-WIS requires labor-intensive configuration, and would benefit greatly to a simplified setup of semantical models

Innovation

Adapt commercially available large language models (LLMs) to transform mnemonic-based data streams, like e.g. WITSML, into a D-WIS style semantic model. 

Value

  • Simplified setup of semantical models accelerates adoption of advanced D-WIS-based automation
  • Including documentation and possibly drawings into the semantical modeling process
  • Generative AI for data aggregation may lead to new data-processing workflows
  • Identification of inconsistencies in D-WIS models, leading to a more robust and user-friendly framework

Status

During 2025 DDBot verified mnemonics to D-WIS conversion with LLM and pointed at possible D-WIS improvements

Next step

Develop multimodal user-assisted and explainable AI solutions to support generation of complex interoperable semantical models for D-WIS

This work is part of the centres workpackage 4 and DigiWells innovation program. 

Contact

image

Sergey Alyaev