SAARTHI

System for Aadhaar Analytics, Risk & Trend Highlighting

SAARTHI is a data-driven analytical framework developed as part of the UIDAI Data Hackathon, aimed at uncovering meaningful societal and administrative insights from anonymised Aadhaar enrolment and update datasets.
The project introduces a novel metric — the Update Dependency Index (UDI) — and applies statistical anomaly detection to identify regions exhibiting abnormal update behaviour.

---

🎯 Problem Statement

Aadhaar enrolment and update activities reflect large-scale demographic, societal, and operational dynamics across India.
While high-level statistics are available, there is limited analytical visibility into post-enrolment update dependency, regional instability, and anomalous update patterns.

SAARTHI addresses this gap by:

• Quantifying update dependency using a unified index
• Detecting statistically abnormal regions
• Translating analytics into actionable insights for governance and system improvement

---

🧠 Key Concepts

Update Dependency Index (UDI)

UDI measures how dependent Aadhaar records are on post-enrolment updates.

UDI = (Total Demographic Updates + Total Biometric Updates) / Total Enrolments

• Low UDI → Stable Aadhaar lifecycle
• High UDI → Frequent corrections or lifecycle-driven updates

---

Anomaly Detection

Statistical Z-score–based anomaly detection is applied to UDI values to flag regions with unusually high update dependency, representing potential risk signals or areas requiring administrative attention.

---

📊 Datasets Used

The analysis uses anonymised, aggregated datasets provided by UIDAI:

1. Aadhaar Enrolment Dataset

   • Age-wise enrolment counts (0–5, 5–17, 18+)
   • Spatial attributes: State, District, Pincode
   • Temporal attribute: Date

2. Aadhaar Demographic Update Dataset

   • Aggregated demographic update activity across age groups and regions

3. Aadhaar Biometric Update Dataset

   • Aggregated biometric update information reflecting revalidation and lifecycle changes

Due to large data volume, datasets are provided as multiple state-wise CSV files and consolidated programmatically.

---

🛠 Methodology Overview

1. Consolidation of state-wise CSV datasets
2. Robust parsing of mixed-format date fields
3. Aggregation of age-wise enrolments and updates
4. Dataset integration using spatial and temporal keys
5. Computation of Update Dependency Index (UDI)
6. Statistical anomaly detection using Z-score analysis
7. Visualisation and interpretation of findings

---

📈 Key Insights

• Most regions exhibit low update dependency, indicating stable Aadhaar lifecycles
• A limited subset of districts and pincodes shows disproportionately high UDI values
• Demographic and biometric updates jointly contribute to observed instability
• Anomalous regions are structurally distinct from normal regions

---

🧪 Technology Stack

• Language: Python
• Libraries:
  • Pandas
  • NumPy
  • SciPy
  • Matplotlib
• Environment: Jupyter Notebook

---

📁 Repository Structure

SAARTHI/
│
├── data/
│ ├── enrolment-data/
│ ├── demographic-data/
│ └── biometric-data/
|
├── saarthi.ipynb
├── SAARTHI.pdf
├── README.md
└── LICENSE

---

🚀 How to Run the Analysis

1. Clone the repository:

   git clone https://github.com/your-username/SAARTHI.git
   cd SAARTHI
   

2. Install required dependencies:

   pip install pandas numpy scipy matplotlib
   

3. Launch Jupyter Notebook:

   jupyter notebook
   

4. Open and run:

   saarthi.ipynb
   

Ensure the dataset folders are placed under the data/ directory as shown above.

---

🔒 Data Privacy Notice

This project uses only anonymised and aggregated datasets provided for the UIDAI Data Hackathon. No personal or sensitive resident-level information is used or inferred.

---

👤 Author

Vrajkumar Shah

B.Tech, Computer Science & Engineering

Dharmsinh Desai University, Nadiad

---

📜 License

This project is released under the MIT License. See the LICENSE file for details.

---

⭐ Acknowledgements

• Unique Identification Authority of India (UIDAI)
• National Informatics Centre (NIC)
• Ministry of Electronics and Information Technology (MeitY)

for providing the datasets and organising the hackathon.

---