A new study published in Engineering presents an adaptive hybrid edge-cloud collaborative offloading method designed to address the challenges of large-scale computational tasks in intelligent machine tools, achieving improved latency, energy efficiency and security performance for complex machining-related computing work. Led by researchers from Jilin University and Beijing University of Technology, the research targets the core pain points of intelligent machine tool operations, where multi-component degradation and dynamic machining task updates generate massive multi-source sensor data and computationally intensive tasks with intricate data dependencies, posing great challenges to traditional single-mode computing frameworks.

The proposed Adaptive Hybrid Edge-Cloud Collaborative Offloading (AH-ECO) mechanism integrates single-edge-cloud and multi-edge-cloud collaboration modes, enabling dynamic switching based on computational node status, task characteristics, dependency complexity and resource availability. The research team constructed a tailored multi-objective optimization model for intelligent machine tools that simultaneously minimizes processing latency, energy consumption and security risks, establishing mathematical models for both single and distributed multi-edge-cloud collaboration modes to quantify latency, energy consumption and security risk metrics respectively. To solve the large-scale task allocation problem in heterogeneous edge-cloud environments, a novel Hybrid Hyper-Heuristic Operator Parallel Evolution (HHOPE) algorithm was developed, which combines genetic algorithms, particle swarm optimization and sparrow search algorithm as core operators, with a multi-feature fusion task pre-assignment mechanism and game-theoretic cross-learning strategy to enhance initialization quality and balance convergence speed and solution diversity.

Extensive numerical and simulation experiments validated the performance of the proposed method against classical and state-of-the-art algorithms. Results showed the AH-ECO mechanism achieved an average 27.36% reduction in task processing time and a 7.89% improvement in energy efficiency compared with advanced techniques, while maintaining superior security performance. A case study on the digital twin gantry five-axis machining center further verified the mechanism’s effectiveness in real manufacturing scenarios covering multi-source concurrent data processing, complex dependency task collaboration, high-computational machine learning workloads and continuous batch task deployment. In this practical validation, the method reduced latency by 37.03% and optimized energy use by 25.93% relative to previous-generation collaboration methods, and in key stages of digital twin machine tool operation, achieved up to 53.02% latency reduction and 29.97% energy consumption optimization.

The research provides theoretical and technical support for sustainable and secure computational offloading in intelligent machine tools, contributing to the development of next-generation smart manufacturing systems. The research team notes that future work will focus on offloading strategies for emerging multi-modal perception tasks of intelligent machine tools and online decision-making methodologies based on deep reinforcement learning to further improve real-time performance of task offloading in dynamic manufacturing environments.

The paper “An Adaptive Hybrid Edge-Cloud Collaborative Offloading Method for Large-Scale Computational Tasks of Intelligent Machine Tool: Low-Latency, Energy-Efficient, and Secure,” is authored by Zhiwen Lin, Kaien Wei, Yiqiao Wang, Chuanhai Chen, Jinyan Guo, Qiang Cheng, Zhifeng Liu. Full text of the open access paper: https://doi.org/10.1016/j.eng.2025.09.030. For more information about Engineering, visit the website at https://www.sciencedirect.com/journal/engineering.