1. A Condition and Fault Prevention Monitoring System for Industrial Computer Numerical Control Machinery
- Author
-
Mattia Ragnoli, Marino Pavone, Nicola Epicoco, Giordano Pola, Elena De Santis, Gianluca Barile, and Vincenzo Stornelli
- Subjects
Artificial intelligence ,CNC ,digital twin ,fault detection ,Industry 50 ,Internet of Things ,Electrical engineering. Electronics. Nuclear engineering ,TK1-9971 - Abstract
Nowadays, the integration of smart systems within the modern industrial scenario is a continuously growing paradigm. Computer Numerical Control (CNC) machinery can heavily benefit from the introduction of Artificial Intelligence (AI) based monitoring applications. In this paper, we present an industrial condition and fault prevention monitoring system for CNC tools. The developed system is the result of an industrial project aimed at realizing a multi-purpose machine which is currently in pre-commercial stage. The results of this work represent the base platform for the further commercial development, which will be carried on from the industrial partners in accordance with clients feedbacks and specifications. This work presents the hardware architecture of the system, the web-based monitoring platform for remote management, and the AI framework used for fault monitoring. The multi-purpose machine is equipped with accelerometer units to monitor the vibration in multiple points of the structure. The control unit of the machine is connected to the sensing nodes and is used to communicate the actual machine state to a remote web platform. The accelerometric data are analyzed through an AI algorithm to perform fault detection. The fault detection algorithm was trained with the measurements performed on the machine under controlled environment faulty operation. The Internet of Things (IoT) based architecture has proven to be effective to facilitate the supervision of the machining processes, and the AI-based classification shows good classification performances for the fault detection tests.
- Published
- 2024
- Full Text
- View/download PDF