Smart Water Security with AI and Blockchain-Enhanced Digital Twins

A scalable, secure, AI- and blockchain-based digital twin platform for rural water management.

This repository contains the implementation of a secure, intelligent digital twin (DT) platform for rural water distribution networks. The system integrates LoRaWAN-based data acquisition, an AI-powered Intrusion Detection System (IDS) using LSTM Autoencoder and Isolation Forest, and a private Ethereum blockchain with Proof of Authority (PoA) to ensure tamper-proof and transparent water management.

---

🌐 Project Description

Water systems in under-connected rural areas are vulnerable to data tampering, inefficiencies, and delayed responses. This project proposes a modular framework combining:

• Digital Twins for real-time monitoring and forecasting.
• Blockchain for secure data storage and billing automation.
• AI for early leak detection and cyberattack prevention.

🧠 Architecture Overview

Figure 1: A Digital Twin Platform in the Water Industry.

---

🔗 Blockchain Integration

Figure 2: Technologies in the Platform on the Blockchain Side.

---

📜 Smart Contract Design

Figure 3: Proposed Smart Contract Structure for the DT Platform.

---

🚿 Leakage Detection Results

Figure 4: Night consumption heatmap for a water meter with leakage.

Figure 5: Comparison of normal and median usage patterns.

Figure 6: Leaked meters and their consumption during nights.

---

🧪 Anomaly Detection Results via IDS

Figure 7: Comparison between night leakage pattern and injected spikes.

Figure 8: Detected anomalies compared with median usage baseline.

---

🔎 IDS Performance Heatmap

Figure 9: Heatmap of Precision, Recall, and F1-Score per attack type.

---

🔧 Technologies Used

Component	Technology
Sensor Layer	LoRaWAN (Simulated)
IDS	LSTM Autoencoder + Isolation Forest
Blockchain	Ethereum (Private, PoA)
Smart Contracts	Solidity (Water Meter Logging)
Visualization	Grafana, Prometheus
Deployment	Docker, Python, Web3.py

---

📦 Key Features

• 🔐 Hybrid IDS filters spoofed, replayed, and anomalous records.
• 🧠 LSTM Autoencoder detects time-based anomalies.
• 🌲 Isolation Forest flags statistical outliers.
• ⛓️ Smart Contracts manage meter registration and payments.
• 🚿 Leak Detection through night-time pattern analysis.
• ⚡ Real-Time Monitoring with Web3 + Digital Twin UI.
• 💶 Cost-Efficient with < €50/month PoA infrastructure.

---

📁 Directory Structure

.
├── algorithms/ # AI IDS implementation
├── contracts/ # Smart contract (Solidity)
├── data/ # Sample water consumption datasets
├── docker/ # Private Ethereum network setup
├── notebooks/ # IDS training and visualization
├── results/ # Figures and evaluation results
└── README.md

---

📜 Smart Contract Overview

The Solidity smart contract handles:

• registerMeter() for adding new meters
• logWaterData() for secure logging
• calculatePayment() for auto billing
• getWaterLogs() for auditing

struct WaterData {
    uint256 timestamp;
    uint256 waterUsage;
    uint256 errorCode;
    string meterId;
}
mapping(string => WaterData[]) public waterLogs;
function logWaterData(string memory id, uint256 usage, uint256 errCode) public { ... }

---

⚙️ Performance & Scalability

Transaction Latency and Throughput

Batch Size	TPS	Mean Latency (s)	Max Latency (s)
1	110	1.2	2.1
5	96	1.5	2.4
10	89	1.7	2.8
20	81	2.1	3.5

💰 Cost Estimation (6-Month Pilot)

Component	Cost (EUR)	Notes
Hetzner VPS	€120–300	Server hosting for validator nodes
Maintenance	~€50	Occasional updates
Energy	Included	Covered by provider
Smart Contracts	Near-zero	Custom PoA with gas price = 0

---

🤝 Acknowledgments

This project is part of the Spanish C107/23 initiative: "Artificial Intelligence Applied to Cybersecurity in Critical Water and Sanitation Infrastructures" Funded by the European Union – Next Generation EU, coordinated by INCIBE.

---

📚 How to Cite This Work If you use this research in your work, please cite it using one of the following formats:

📌 APA (7th Edition)

arXiv version:

Homaei, M., Gonzalez Morales, V., Mogollón Gutiérrez, Ó., Molano Gómez, R., & Caro, A. (2025). Smart Water Security with AI and Blockchain-Enhanced Digital Twins. arXiv. https://doi.org/10.48550/arXiv.2504.20275

---

📌 IEEE

arXiv version:

M. Homaei, V. Gonzalez Morales, Ó. Mogollón Gutiérrez, R. Molano Gómez, and A. Caro, “Smart Water Security with AI and Blockchain-Enhanced Digital Twins,” arXiv, 2025. [Online]. Available: https://doi.org/10.48550/arXiv.2504.20275

---

📌 Chicago (Author-Date)

arXiv version:

Homaei, Mohammadhossein, Victor Gonzalez Morales, Óscar Mogollón Gutiérrez, Ruben Molano Gómez, and Andres Caro. 2025. Smart Water Security with AI and Blockchain-Enhanced Digital Twins. arXiv. https://doi.org/10.48550/arXiv.2504.20275

---

📌 BibTeX

@misc{homaei2025smart,
  author       = {Homaei, Mohammadhossein and Gonzalez Morales, Victor and Mogollon Gutierrez, Oscar and Molano Gomez, Ruben and Caro, Andres},
  title        = {Smart Water Security with AI and Blockchain-Enhanced Digital Twins},
  year         = {2025},
  publisher    = {arXiv},
  doi          = {10.48550/arXiv.2504.20275},
  url          = {https://arxiv.org/abs/2504.20275},
  note         = {arXiv preprint arXiv:2504.20275},
  keywords     = {Cryptography and Security (cs.CR), Artificial Intelligence (cs.AI), Machine Learning (cs.LG)}
}

Contributors

• Hubert Homaei, Victor Gonzalez, Oscar Mogollon.

---

License

This project is licensed under the GPL-3.0 License - see the LICENSE file for details.