# A decentralized blockchain-based smart framework for continuous vehicle emission monitoring in smart cities

## TL;DR

This paper introduces a blockchain-based system using IoT and machine learning to monitor and reduce vehicle emissions in smart cities.

## Contribution

A novel blockchain-integrated framework with XGBoost ML for real-time, tamper-proof vehicle emission monitoring and prediction.

## Key findings

- The blockchain-based XGBoost model achieved 97.98% prediction accuracy, outperforming cloud systems by 4.9%.
- The system delivered a throughput of 679 Mbps, response time of 91.98 milliseconds, and processing time of 225.88 milliseconds.
- The framework addresses PUCC system limitations with transparent, decentralized emission monitoring.

## Abstract

Due to its high emissions, vehicular air pollution remains a critical contributor to environmental degradation and global warming; however, even in smart cities, control mechanisms often remain inadequate. The current Pollution Under Control Certificate (PUCC) system suffers from inefficiencies such as weak monitoring, maintenance gaps, and data manipulation risks. This paper proposes a blockchain-enabled framework integrating Internet of Things (IoT) sensors, machine learning (ML), and decentralized data validation to enhance emission control. In the proposed system, IoT-based sensors installed in vehicles continuously monitor emission levels and transmit real-time data to a blockchain network, ensuring tamper-proof, transparent, and immutable records. A consortium blockchain is used to validate and store emission data across distributed nodes. Furthermore, the eXtreme Gradient Boosting (XGBoost) machine learning model is applied to this data to predict emission trends and identify vehicles requiring maintenance proactively. Comparative simulations with cloud and fog-based models demonstrate the system’s superiority: the blockchain-based XGBoost model achieved 97.98% prediction accuracy, outperforming cloud systems by 4.9%. Additionally, the proposed system delivered a throughput of 679 Mbps, the response time of 91.98 milliseconds, and a processing time of 225.88 milliseconds. This framework overcomes PUCC system limitations, offering a scalable and reliable approach for reducing vehicular pollution in support of smart cities and sustainable urban development.