Isatis.py: Eramet’s strategic choice to standardize, secure, and industrialize resource estimation across its deposits

(Cliquez ici pour la version française)

Eramet is a leading international mining player, specialized in the extraction and processing of critical metals for the energy transition (manganese, nickel, lithium, etc.). Present in more than 20 countries, the group relies on innovation to optimize its processes and contribute to the development of a more sustainable mining industry.

THE CHALLENGE

Before adopting the Isatis.py Python geostatistics library, resource estimation at Eramet relied on a legacy of Isatis batch processes, later Isatis.neo, and various Python scripts often customized by each geologist.

This approach had several limitations:

• Complex maintenance, requiring experts with combined skills in geosciences, Python, and knowledge of Isatis.neo.
• Lack of uniformity, with estimates varying from one site to another and from one user to another.
• Weak traceability, with little historical tracking of inputs and outputs related to calculations.
• Risk of errors due to manual modifications of scripts by users.
• High reparameterization times during model updates, particularly in the production phase.

➡️ The ambition was clear: to deeply rethink the process to ensure standardization, reliability, and transparency of resource estimation workflows, while guaranteeing their scalability, regardless of the deposit type. ⬅️

THE SOLUTION

After a phase of comparative testing of different Python libraries, including both open-source and proprietary solutions, Eramet chose Isatis.py.

Why Isatis.py? Four key reasons:

• Performance – Comparable computation times to Isatis.neo and much faster than many other solutions.
• Reliability – High-quality results, aligned with the group’s geostatistical standards, validated through cross-checking.
• Documentation and support – Clear documentation and highly responsive technical support.
• Geovariances’ expertise – A world-renowned reputation in geostatistics.

Implementation:

The library is encapsulated within an internal platform with a web interface. Geologists no longer handle code directly; they only enter the key parameters (such as variograms and neighborhood). The entire estimation logic is industrialized within a configurable and automated workflow, with automatic generation of figures and analysis files. The core of the process remains invisible, but fully controlled.

THE BENEFITS

• One single codebase, dozens of deposits, estimated consistently
  Standardization ensures total consistency of results and strengthens data quality for planning. 
• Drastic reduction of human errors
  The final user does not handle scripts, focusing only on relevant geostatistical parameters.
• Speed and agility
  Estimates are relaunched within minutes. Updates (e.g., infield or grade control) become virtually instantaneous.
• Resilience to turnover
  Integrated documentation, stored histories, and process unification ensure operational continuity.
• Interoperability with Isatis.neo
  All intermediate files can be exported for advanced analysis or visualization in Isatis.neo.
• Overall improvement of the process
  Structuring the workflow standardizes data formats, clarifies methodological choices, and brings greater rigor across the processing chain.
  

CONCLUSION

➡️ Isatis.py is more than just a geostatistics library. It is the cornerstone of a group-wide standardization strategy. ⬅️

By centralizing estimations around a single, configurable, and documented script, Eramet drastically reduces the risk of human error and eliminates methodological inconsistencies. Each run maintains a complete history of parameters, input data, and generated results, ensuring rigorous traceability and transparency. This approach ensures homogeneous, reliable, and auditable estimates while accelerating updates in production contexts.

The goal is set: to generalize the use of standardized workflows for a unified and controlled resource management.