Introducing MineProductivity v2.0 - A Stable Foundation for Mining Productivity Intelligence

[imanojkumar]

Introducing MineProductivity

I'm pleased to announce the first stable release of MineProductivity.

After months of designing, implementing, testing, documenting, and validating the architecture, the framework has reached v2.0.

MineProductivity is an open-source Python framework for mining productivity intelligence, providing a governed foundation for KPIs, analytics, digital twins, simulation, optimization, AI agents, and visualization.

The framework is organized as fourteen packages along a strict dependency chain:

core → ontology → events → kpis → analytics → decision
→ digital_twin → simulation → optimization → agents → visualization

plus three cross-cutting infrastructure packages (registry, plugins, and connectors) that any layer may use. Every performance indicator is a discoverable, versioned, self-describing object rather than a formula embedded in a script or a spreadsheet cell.

Engineering quality and production readiness

The v2.0 release was gated on a recorded certification pass. At certification:

• 2,986 automated tests pass across the full suite.
• mypy --strict is clean across 314 source files.
• Branch coverage is 99%; the small residual is interface-only abstract methods and documented concurrency guards, disclosed in the source.
• ruff lint and format are clean; documentation builds with zero broken links.
• Cross-platform CI, CodeQL analysis, and dependency review run on every change.

An invitation to contribute and collaborate

The architecture is complete and locked, which makes this a good moment to build on the framework. The most valuable contributions now are at the extension points:

• concrete solver adapters for the optimization paradigms;
• reasoning backends and tools for the agent layer;
• rendering backends for visualization;
• connector adapters for real fleet-management and dispatch systems;
• additional KPIs and governed domain artifacts;
• and applications built on top - a CLI, dashboards, and domain studies.

Contributions of documentation, examples, benchmarks, and real-world validation are equally welcome. Mining practitioners, researchers, data scientists, and software engineers each see a different part of this problem clearly; the framework is designed so those perspectives can be added as plugins and data rather than as forks. Please read the contributing guide and the engineering rules before opening a pull request, and feel free to start a discussion if you are considering a larger piece of work.

Links

• GitHub: https://github.com/imanojkumar/MineProductivity
• Documentation: https://imanojkumar.github.io/MineProductivity/
• PyPI: https://pypi.org/project/mineproductivity/ - pip install mineproductivity
• Zenodo (Concept DOI): https://doi.org/10.5281/zenodo.21172767

MineProductivity is released under the Apache License 2.0.