WP4 Description

The main goal of WP4 is to use machine learning and data science to improve the modelling and visualization of complex data. It is split into 4 main tasks. WP4 is tightly linked to all the other WPs in an iterative feedback loop process. 

Task 4.1:  develop innovative open-source 3D modelling methods based on geology, geochemistry and geophysics 

The goal of this task is to create fully reproducible scripts to generate 3D quantitative models of the subsurface consistent with as much data as possible for each VECTOR case study. These models include the 3D geometry of the most relevant geological formations as well as geophysical, petrophysical and petrochemical distributions within each unit. To do so, we will contribute to and develop open source software packages. We will use machine learning methods to process the raw data into information useful for the construction of digital models.  These models will be compared with petrophysical data to better understand possible relations between chemistry/rock, composition/mineralogy, and petrophysical properties. Resulting relationships will then be utilized to derive vectors toward mineralized zones as well as defining the uncertainties associated with these vectors. 

Task 4.2: produce evolutive mineral deposits models to better define mineralization pathways

The new relationships defined within Task 4.1 will be used in construction of new 3D and potentially 4D models of the geology at the local (deposit) to regional (sub-basin) scale. We will simulate the thermal and mechanical history as part of the basin evolution using a geological finite element model. We will then develop basin systems models and apply geochemical and reactive transport characteristics in order to represent mineral systems processes. This will serve fit-for-purpose geochemical models and process descriptions to allow a deliver an accurate distribution of predicted mineral deposits.

Task 4.3:  create interactive maps displaying the main parameters that affect shared value

Task 4.3 will bridge the interdisciplinary fields of remote sensing, deep learning, and social sciences to identify and map the spatial distribution of key “vectors” or indicators that play a role in shared value at regional scale. The advantage is that the analysis moves away from a site-specific focus to a regional scale, and is closely aligned with the decision-making tools required to balance competing interests under the rubric of the United Nations Resource Management system (UNRMS).

We will compile and evaluate available geospatial data including land use, demographic, environmental, remote sensing, and economic data, plus spatially located data including indicators of socio-demographics, attitudes, values, governance mechanisms, history of mining activities, conflict history, and policy drivers and develop an innovative methods to automatically source useful and relevant information from open-access scientific and social science articles specifically related to mining projects. For that we will create a deep learning architecture to analyze heterogenous data to identify and assess the spatial dynamics of key parameters that specifically relate to shared value. Results will be incorporated into the revised SVi in Task 3.2.

Additionally we will develop the backend of an interactive tool that visualizes social, environmental and governance factors which are critical for the accessibility of CRMs in Europe. These outputs will be fed back to WP3 for iterative quality review and validation and for review by relevant stakeholders and incorporation into an interactive tool which combines the identification of geological and social vectors to increase the accessibility of CRMs in Europe in Task 4.3. This task will be monitored closely by the independent ethics committee to ensure consensual and ethical development. 

Task 4.4:  generate the backbone of an interactive platform for the visualization of integrated geodata and social data to support decision making and education by policy makers, industry, and the public.

The results from WP4 will be presented in a platform that integrates spatially referenced geological and social vectors that are critical to critically discuss the shared value around CRMs in Europe. This platform will support evidence based decision making about CRMs and will be able to be used by all stakeholders including governments, relevant industries, researchers, NGOs, and the public as they consider the supply chain for the CRMs needed to meet the demands of the EU Green Deal. This platform will be created in an iterative process whereby experts and users assess and evaluate it, and their feedback will be used for refinement and enhancement. We will develop a platform that highlights where new resources of CRMs are likely to occur, combined with where the development of these resources has a fair chance to define shared value. This tool will also deliver wider benefits in accordance with the UNRMS and UN SDGs. We will develop the interface to visualize, edit and communicate the resulting models in intuitive 3D with real-time collaboration across different devices. This will include support for virtual reality, augmented reality, as well as common desktop operating systems. For 3D visualization, we will use LiquidEarth, a cloud platform being developed by partner Terranigma.