BekzatK/shyrai-a1c-quality-control

Shyrai A1c Quality Control Dataset

Description

This dataset contains real-world quality control (QC) data collected during the production and testing of the Shyrai A1c glycated hemoglobin analyzer, used for diabetes diagnostics.

The dataset is designed to support research in:

• —AI-driven quality management systems (QMS)
• —anomaly detection
• —predictive quality analytics
• —medical device manufacturing

Dataset Structure

The dataset includes the following types of data:

• —Production data: lot ID, production date, serial numbers
• —Measurement data: repeated measurements, mean values
• —Reference values: ABX Pentra C400
• —Quality metrics:
• —Bias (%)
• —CV (%)
• —Pass/Fail status
• —Test conditions:
• —temperature (10°C, 40°C)
• —control material levels
• —Biological samples:
• —venous blood
• —capillary blood
• —Error logs:
• —error codes (101–310)
• —error categories
• —Derived features:
• —ML-ready features
• —IDQI-HbA1c quality index

Use Cases

This dataset can be used for:

• —quality classification (Pass/Fail)
• —anomaly detection in production
• —prediction of measurement deviations
• —development of digital twin models
• —AI integration in QMS systems

Data Format

The dataset is provided in CSV format, structured for machine learning (ML-ready).

Source

Data collected from real production processes at:

LLP “Axel and A” — the first domestic manufacturer of diagnostic test strips for glucose, cholesterol, and triglyceride measurement.

Limitations

• —small dataset size (pilot study)
• —limited to one analyzer type (Shyrai A1c)
• —further validation required for generalization

Citation

If you use this dataset, please cite:

Karimkyzy, B. (2026). Shyrai A1c Quality Control Dataset for AI-driven QMS research.

Author

Bekzat Karimkyzy — PhD doctoral student in Information Technologies (professional track), D. Serikbayev East Kazakhstan Technical University; employee of Axel & A LLP. ORCID: 0009-0004-4696-7927

Future Work

• —dataset expansion
• —integration with real-time monitoring systems
• —development of predictive AI models