Toward digital twins for intelligence manufacturing: Self-adaptive control in assisted equipment through multi-sensor fusion smart tool real-time machine condition monitoring

Compared to traditional monitoring methods, multi-sensor fusion smart tool offers several advantages, including full-process monitoring and a broader range of applications (e.g., flat, curved, and complex surfaces). When integrated with artificial intelligence models for tool state monitoring, these tools provide strong generalization capabilities and high prediction accuracy. They can also adjust machine tool process parameters to extend tool life. However, the quasi-in-situ regulation of cutting parameters has a limited scope, making it challenging to achieve full working condition adaptability. The introduction of assisted equipment can enhance process adaptability. Furthermore, adaptive control mechanisms can regulate the machining process to reduce energy consumption by adjusting the opening and closing parameters. Despite these advantages, the linkage control mechanism for the smart tool remains unclear, and existing tool wear models struggle to adapt to variable working conditions across multiple scenarios. To address these challenges, this paper explores the digital twin modeling and application of smart tool machining processes. First, a digital twin-driven tool machining process model is developed, with an exploration of specific application scenarios and methods. Secondly, an adaptive coupling mechanism for assisted equipment based on digital twins is established, which simultaneously improves machining quality and reduces energy consumption. Additionally, the online tool wear identification model is enhanced to increase its generalization and reduce the cost of model reconstruction when working conditions change, thus enabling green intelligent manufacturing under high-quality machining conditions.