Geological Assistant

The Geological Assistant is a SIRIUS innovation project between university researchers, Equinor, and Schlumberger.


The project’s goal is to develop a tool-supported method for exploration geologists to better assess and evaluate exploration prospects by applying established techniques from knowledge representation and formal methods from software verification. The project includes researchers from the University of Oslo and NTNU with expertise ranging from implementation and use of digital technologies, knowledge representation, formal methods, and naturally, geology.

Geological Multi-scenario reasoner (GeMS) 

Interpretation of the subsurface in order to find out where hydrocarbons are located is a challenging task for explorationists. They need to be creative and come up with innovative ideas when defining and assessing new prospects, especially nowadays when the easy to find, big fields have been already discovered. The challenges related to prospect assessment are

      1. geodata is uncertain, intermittent, sparse, multiresolution, and multi-scale, and
      2. explorationists often limit themselves to assess few possible scenarios  


Recent advancements in computation, network and storage have led to numerous opportunities to improve these subsurface evaluation workflows. Further, the volatility and uncertainty in the oil and gas industry have forced exploration and production companies to find improved and cost-effective solutions by automating thesse workflows. 


When it comes to digitalization, traditionally, the focus has been on purely data-driven workflows. Although geological reasoning is the most crucial factor that defines exploration success, little attention has been given to exploit digitalization opportunities in reasoning-based evaluation. In geological reasoning workflows, explorationist still rely on ad hoc manual work practices and tools and use pen and paper along with computer drawing and presentation tools to develop and communicate multiple hypothetical geological scenarios of the prospect. This leaves them with little to no efficient means to make the fullest use of state-of-the-art digital technologies to communicate and systematically compare and assess different hypothetical geological scenarios before deciding which scenario to pursue when assessing exploration prospects.

A recent study * of 97 wells drilled in the UK sector of the North Sea from 2003 to 2013 showed that some of the major reasons for unsuccessful exploration were the fact that several of the prospects relied too heavily only on data (seismic DHIs and amplitude). Regional play-based work for setting and context were repeatedly missed, and pre-drill analysis was often excluded the full range of possible outcomes.

The industry’s current trend on moving from a physics-driven world to a purely data-driven world for a complex domain like Geoscience has proved inefficient. We believe that a significant success factor will be to combine geological reasoning- based evaluation with the insights derived from the Geological, Geophysical, and Petrophysical data.

* C.Mathieu.MorayFirth–centralnorthseapostwellanalysis.Oil&GasAuthority,2015.

Our Approach

The geological multi-scenario reasoning methodology developed in this project provides a hybrid approach by combining the data-driven seismic interpretation (faults and horizons) with geological reasoning (based on the encoded geological rules). It can significantly help subsurface experts to think out of the box to consider several geological models rather than relying on a single model. Further, it will enable Geoscientists to consider a full range of possible scenarios and corresponding outcomes beyond what is possible within human capacity.

We have experimented with logic- based techniques for subsurface modeling, with focus on how the inherent complexity in geology such as spatial and temporal aspects can be formally captured and reasoned about using the strength of different formalizations. In particular, we demonstrated the use of abstraction and how formal modelling gives a precise and human-readable representation of domain knowledge. Further, we developed a mechanism to bring together the various models in a novel tool-based approach that constructs multi-scenarios to support geologically oriented subsurface evaluation. In this work, we combined techniques from knowledge representation with formal methods (mathematical approaches that support the rigorous specification, design and verification of computer systems) to the exploration domain.

This logic-based technology enables explorationists to express interpretive uncertainty as discrete scenarios with branches of potential alternative interpretations. With this approach, common-sense explorationist domain knowledge and rules of thumb are explicitly represented in the tools together with collected O&G data.


Demo Videos

[Coming Soon]


Posters [UseCase, Technology]
Documentation for each module

Can be found in the different repositories containing the source code

Source Code

The source code of SiriusGeoAssistant is  available on Gitlab under the Apache License Version 2

Proto-Scenario Generators and Tools [GitLab]

Proto-Scenario generator for the SIS usecase written in Python. Proto-scenario generator to create Maude-files from an OWL-ontology for use with the JS usecase. Experimental tool to convert data from a Petrel-export into an OWL-ontology.

Scenario Generation / Simulation [GitLab]

Maude-programs to turn proto-scenarios into scenarios and run simulations on those for both use cases.For reference: Maude 3 source code and documentation.

Toolchain  [Gitlab]

Java application that integrates proto-scenario generation and scenario creation and simulation into one toolchain accessed as webservices and which orchestrates the execution for both use cases.

Querying [GitLab]

An experimental querying tool written in Python to navigate the large result-sets of the simulation.


Ingrid Chieh Yu, Irina Pene, Crystal Chang Din, Leif Harald Karlsen, Chi Mai Nguyen, Oliver Stahl, Adnan Latif. Subsurface Evaluation through Multi-Scenario Reasoning. In Interactive Data Processing and 3D Visualization of The Solid Earth, Springer (2022).

Crystal Chang Din, Leif Harald Karlsen, Irina Pene, Oliver Stahl, Ingrid Chieh Yu, and ThomasØsterlie. Geological multi-scenario reasoning. NIK: Norsk Informatikkonferanse (2019).

Vegar Skaret. Knowledge Representation and Concretization of Underdetermined Data. Master’s Thesis (2020).

Østerlie, Thomas, Elena Parmiggiani, and Eric Monteiro (2017). Information infrastructure in the face of irreducible uncertainty. The 5th Innovation in Information Infrastructure (III) workshop, November 7-9, Rome, Italy.

Monteiro, Eric, Thomas Østerlie, Elena Parmiggiani, and Marius Mikalsen (2018). Quantifying quality: Towards a post-humanist perspective on sensemaking. In Aanestad, Margunn, Magnus Mӓhring, Carsten Østerlund, Kai Riemer, and Ulrike Schultze (Eds.), in Living with Monsters? Social Implications of Algorithmic Phenomena, Hybrid Agency, and the Performativity of Technology (pp. 48-63), Cham, Switzerland: Springer International Publishing.

Østerlie, Thomas, Elena Parmiggiani, and Petter Almklov (2019). History-based geological modelling: Some elements of a design theory. The 6th Innovation in Information Infrastructure (III) workshop, September 18-20, Guildford, UK



Thomas Østerlie, Ingrid Chieh Yu , Crystal Chang Din,   Jens Otten,  Irina Pene,  Michael Heeremans,  Adnan Latif [Contact Person], Leif Harald Karlsen, Fabricio Rodrigues, Vegar Skaret, Eric Monterio and Elena Parmiggiani



Hallgrim Ludvigsen



The Geo-Assistant team would like to thank project partners Equinor, Schlumberger and NTNU for the discussions and feedback on this work.

This work was supported by the SIRIUS Centre for Scalable Data Access (Research Council of Norway, project 237889).