Capability and Capacity Statement
Objective
To highlight Datacode’s experience, capabilities and capacity with respect to Geological -Data management solution for any AI/ML Geo-scientific projects in the space of green-field mineral prospecting.
Capability of Datacode
Team Datacode, Has experience and knowledge to carry out –
- Geological, Geophysical and Geochemical mapping, data processing,
interpretation and modeling - Resource modeling and estimation
- Creating and ensuring quality of data for ML and AI project
- GIS Data management and Analysis
- QA/QC of geophysical data and statistical analysis or Geochemical data
- Work in different terrain to ensure the field objective of data collection is properly
performed - Defined SOP’s for various mineral exploration task
Capacity of Datacode
Team Datacode consists of geoscientists who understand the importance and utility of geoscientific data beyond just numbers, but their significance is with respect to mineral exploration and mineral systems. The team understands the process from data to prospectivity to resources modeling and more. We have;
1. Team of 18 Geologists, 5 Geophysicists who have been involved in various
projects to ensure
- Smooth transition from data wrangling ➔ actionable insight of the data being captured/audited, to ensure free-up in-house geoscientist to focus
on geological interpretation or higher order tasks. - Ensure a single data source, which can be effectively used across multiple technologies.
- Understand the workflow and ensure multiple options for constrained
modeling of a prospective block. - Statistical analysis of data leading to optimization of process and modeling
is key to all our process.
- Smooth transition from data wrangling ➔ actionable insight of the data being captured/audited, to ensure free-up in-house geoscientist to focus
2. Having access to university, industry trained geoscientist and existing relationship similar service providers Datacode, has ability to upscale the capacity at a pretty short notice. In addition to this, Datacode has easy access to IT savvy geoscientist from Indian market and Indian universities. Backed up with a well – defined induction training program to create next generation geoscience data engineers.
3. Knowledge of Python and other development tools.
4. Being resellers of Oasis Montaj™(Geosoft), Leapfrog, Rockworks, Idrisi, Fine software, Mira Geoscience, etc our team has been trained and providing training
to clients on these software’s. This gives them an inside of data utilization by these software’s. They have been working on these software’s for more than 25 years.
5. Proactively working with clients on outcomes of the data while handling/auditing
process.
6. A team of English speaking geoscientists, working across the time zone, young with eagerness to learn and perform.
Background
The key aspect of any AI/ML tools in mineral prospecting is training data sets and constraint modelling. It has been observed that;
- Data captured from historical reports doesn’t contain Metadata. However, metadata is key to understanding the baseline data captured from legacy maps or reports. If the metadata is not captured while capturing the basic data like geology, geochemistry or geophysical data, it is more likely to have erroneous output post application of AI/ML tools.
- The involvement of domain experts (Geologist/Geochemist/Geophysicist) is key to understanding the data. Though the technology is developed by IT experts, the domain knowledge for the matter of understanding/interpretation of data is missing at times.
- Mineral Systems, which is a key aspect leading to development of a prospective mineral deposit. It is one of the key foundation to exploration of deep-seated mineral deposits which need’s to be incorporated.
- Though a volume of data is available in geoscience industry, however variation of the data is just limited by the process. Geoscience data never dies, being non- transactional data, it can be used over decades if captured and audited with full confidence.
- Kreuzer* reports (2010) literature review reports a commonly cited “consensus” of 0.5% to 1% success for greenfield and about 5% for brownfield, and failures not being documented properly/or published. It is important to incorporate the reasons of failure in AI/ML models so as to enable for outputs to be properly classified.
*(Kreuzer, O.P. (May 2010). Risk and Uncertainty in Mineral Exploration: Implications for Valuing Mineral Exploration Properties. AIG News No. 100. (Cites Hronsky, 2004; discusses low base-rate and published success-rate ranges.) Classification on these failures, is critical as many times “Exploration organizations often define “failure” solely as the absence of an economic discovery” whereas same deposit can become economically viable over a time, given changes in cost structures or changes to commodity prices. So, in fact such findings should be classified as inventory of prospective blocks.
The Datacode team has developed and designed the process to address the said challenges and more. We at Datacode believe Metadata, data audit(s), capture of failures, classification of failures and domain knowledge of Data capture team are the key components to success of any AI/ML project in the field of greenfield mineral exploration.
Experience
Datacode has successfully delivered large-scale geoscience data management, QA/QC, mineral exploration, and geospatial projects for government agencies and global organizations.
Geological Survey of India (GSI)
Designed and implemented the national geophysical and geochemical database, including database architecture, metadata management, data migration, and loading of multi-terabyte datasets generated over decades.
Performed QA/QC for the National Airborne Geophysical Mapping Programme (NAGPM), covering 911,626 line km (248,625 km²) of magnetic and spectrometric survey data.
Executed QA/QC of TDEM heliborne geophysical data covering 135,422 km² and approximately 50,000 line km of survey.
Seequent
Provided drilling data annotation and QA/QC for the Imago AI Project, supporting machine learning models for automated structural feature recognition from drillhole imagery through continuous collaboration with Seequent’s technical team.
Royal Commission for AlUla (RCU), Saudi Arabia
Developed and populated a comprehensive geo-environmental database for industrial minerals, including structured data management and integration.
Directorate of Mines, Odisha (OMC Project)
Delivered a 5-year Mineral Resource Mapping (MRM) Project, including:
Inventory creation for prospective and auction-ready mineral blocks (UNFC G2 & G4).
Development of the Mineral Atlas Portal.
Digitization of legacy geological data.
Data gap analysis and exploration planning.
Processing, interpretation, and modelling of airborne and heliborne geophysical datasets to identify prospective mineral targets.
MinEx CRC, Australia
Developed a metadata management platform to capture and manage research data across multiple mineral exploration projects, ensuring standardized documentation throughout the project lifecycle.
University of Western Australia & MinEx CRC (LOOP3D Project)
Contributed to the development of automated workflows for extracting lithological and stratigraphic information from drillhole databases, supporting advanced 3D geological modelling and uncertainty analysis for the open-source LOOP3D platform.