Intelligent Geochemistry · Open Science · Next-Generation Isotope Analysis
A cyber-physical framework for intelligent and automated geochemical laboratories
iGeochem Cloud is a comprehensive platform designed to revolutionize geochemical research through artificial intelligence and automation. Based on the article "iGeochem cloud: A cyber-physical framework for intelligent and automated geochemical laboratories", this project aims to build the future of autonomous geochemical research.
- Intelligent Quality Control: Machine learning models with 99.6% F1-score accuracy for anomaly detection
- Automated Data Processing: End-to-end pipeline from raw data to publication-ready results
- Natural Language Interface: Interact with your data using natural language queries
- Self-Driving Mercury Isotope Analysis: Automated pipeline for MC-ICP-MS mercury isotope analysis
- Real-time Quality Monitoring: Instant detection and diagnosis of measurement anomalies
- Root-Cause Analysis: Multi-class classification to identify probable causes of abnormalities
- Findable: Persistent identifiers (IGSN, ORCID, DOI)
- Accessible: Standardized API access
- Interoperable: Ontology mapping (QUDT, ENVO)
- Reusable: Comprehensive metadata and provenance tracking
- Digital Twins: Virtual representations of instruments and processes
- Instrument Abstraction: Unified SDK for multiple instrument brands
- Real-time Monitoring: Lab safety and equipment health tracking
- Distributed Research: Multiple AI agents working together
- Collective Intelligence: Consensus-based decision making
- Specialized Agents: Planner, Researcher, Analyst, Critic, Executor
- Hypothesis Generation: AI-driven scientific hypothesis creation
- Experiment Design: Automated experimental protocol optimization
- Feedback Loop: Continuous learning and improvement
igeochem-cloud/
├── src/
│ ├── core/ # Core infrastructure
│ ├── flagship/ # Flagship projects (Hg-MC-Auto)
│ ├── orchestration/ # AI orchestration layer
│ ├── fabric/ # FAIR data fabric
│ ├── physical/ # Physical layer abstraction
│ ├── user_interface/ # APIs and web interfaces
│ ├── llm/ # Geo-LLM module
│ ├── autonomous_lab/ # Autonomous laboratory core
│ ├── multi_agent/ # Multi-agent system
│ ├── domain_adapters/ # Domain-specific adapters
│ ├── knowledge_graph/ # Knowledge graph
│ ├── simulation/ # Simulation module
│ └── ethics_safety/ # Ethics and safety
├── tests/ # Comprehensive test suite
├── docs/ # Documentation
├── config/ # Configuration files
├── data/ # Sample data
└── scripts/ # Utility scripts
# Clone the repository
git clone https://github.com/IGeochemCloud/igeochem-cloud.git
cd igeochem-cloud
# Create virtual environment
python -m venv venv
source venv/bin/activate # On Windows: venv\Scripts\activate
# Install dependencies
pip install -r requirements/base.txt
pip install -r requirements/ml.txt
pip install -r requirements/llm.txt
# Configure environment
cp config/base.yaml config/development.yaml
# Edit config/development.yaml with your settingsfrom src.flagship.hg_mc_auto import HgMCAuto
# Initialize the pipeline
pipeline = HgMCAuto(model_path="models/hg_mc_auto_v1")
# Process raw data
results = pipeline.process(
raw_data="data/samples/mercury/sample_data.csv",
quality_threshold=0.85
)
# Generate report
report = pipeline.generate_report(results)
print(report)uvicorn src.user_interface.api.main:app --reload --host 0.0.0.0 --port 8000Visit http://localhost:8000/docs for interactive API documentation.
| Metric | Binary Classification | Multi-class Diagnosis |
|---|---|---|
| Accuracy | 99.61% | 99.84% |
| F1-Score | 0.9960 | 0.9909 |
| AUC | 0.999–1.0 | — |
Based on validation with 26,218 historical measurements.
- Python 3.10+: Primary programming language
- PyTorch 2.0+: Deep learning framework
- FastAPI: High-performance API framework
- PostgreSQL: Primary database
- Redis: Caching and message queue
- Transformers: Large language models
- Scikit-learn: Traditional ML algorithms
- LangChain: LLM application framework
- XGBoost/LightGBM: Gradient boosting
- RDFlib: Semantic web and linked data
- Pydantic: Data validation
- SQLAlchemy: ORM framework
- Docker: Containerization
- Kubernetes: Orchestration
- Prometheus: Monitoring
- Grafana: Visualization
We welcome contributions from the community! Please see our Contributing Guide for details.
# Install development dependencies
pip install -r requirements/dev.txt
# Run tests
pytest tests/ -v
# Format code
black src/ tests/
isort src/ tests/
# Type checking
mypy src/If you use iGeochem Cloud in your research, please cite:
@article{huang2026igeochem,
title={iGeochem Cloud: A Cyber-Physical Framework for Intelligent and Automated Geochemical Laboratories},
author={Huang, Qiang and Zhou, Chufan and Feng, Xinbin and Tang, Yang and Zhong, Ying},
journal={Artificial Intelligence in Geosciences},
year={2026},
doi={10.1016/j.aiig.2026.100257}
}Research Group of Qiang Huang
State Key Laboratory of Environmental Geochemistry
Institute of Geochemistry, Chinese Academy of Sciences (IG-CAS)
- Principal Investigator: Dr. Qiang Huang
- Website: IGeochemCloud
- Email: igeocloud@hotmail.com
- Issues: GitHub Issues
- Discussions: GitHub Discussions
This project is licensed under the MIT License - see the LICENSE file for details.
- State Key Laboratory of Environmental Geochemistry, IG-CAS
- National Natural Science Foundation of China
- All contributors and collaborators
Building the Future of Intelligent Geochemistry 🌟