New research published in engineering We present a new framework that combines machine learning (ML) and blockchain technology (BT) to enhance computational security in engineering. The framework named Machine Learning in Blockchain (MLOB) is intended to address the limitations of existing ML-BT integration solutions that focus primarily on data security, while overlooking computational security.
ML is widely used in engineering to solve complex problems, providing high accuracy and efficiency. However, they face security threats such as data tampering and logic corruption. BT has characteristics of decentralization, transparency and immutable and is being investigated to protect engineering data. However, because ML models often run outside of blockchain, traditional ML processes remain vulnerable to off-chain risk.
The MLOB framework places both data and computational processes on the blockchain, running as smart contracts, and protects execution records. It consists of four core components. ML acquisition, ML models are trained for specific tasks. ML transformation adapts trained models for blockchain deployment. ML secure load, data security and model transfer guarantee. Consensus-based ML model implementation. Ensures the safety and accuracy of calculations.
To illustrate the effectiveness of the MLOB framework, researchers developed a prototype and applied it to progress monitoring tasks in indoor construction. They compared the MLOB framework with three baselines and two recent ML-BT integration approaches. The results showed that the MLOB framework significantly enhances security and successfully defends against six designed attack scenarios. Maintaining the highest accuracy, there was only a 0.001 difference in the mean intersection of Union (MIOU) metrics compared to the best baseline method. Although efficiency was slightly compromised, there was a latency increase of 0.231 seconds compared to the most efficient baseline, overall performance met the requirements of industrial practices.
The MLOB framework also has administrative implications. This encourages organizations to innovate by integrating advanced technologies that could lead to more competitive engineering businesses. It also helps reduce risks related to data and logic security, optimize resource allocation, and increase economic resilience.
However, the framework has some limitations. Support for latency sensitive scenarios is limited and there is no user-friendly interface. Future research will focus on optimizing its efficiency, designing more accessible user interfaces, further improving its usability and expanding the application of engineering computing.
Zhiming Dong, Weisheng Lu, published by Weisheng Lu, “Machine Learning (MLOB) in Blockchain: A New Paradigm for Computational Security in Engineering.” Full Open Access Paper: https://doi.org/10.1016/j.eng.2024.11.026. For more information about engineeringplease follow x (https://twitter.com/engineeringjrnl) & Facebook (https://www.facebook.com/engineeringjrnl).
