In industrial contexts, pulley-belt systems are widely used to precisely drive the working head of CNC machines like laser, plasma, and water cutting machines. In high-precision CNC machines, any obstruction in the movement of the working head results in a failed task and can eventually lead to damage to the mechanical structure of the CNC machine.
We will develop a predictive maintenance solution that will use a machine learning model to analyze vibration data collected in the near vicinity of the mounting point of a Timing belt used in a 150W CO2 laser cutting machine and detect possible belt looseness or pulley misalignment before they reach a critical stage in which they might endanger the integrity of the machine.
Predictive maintenance is a very good alternative not only to minimize the presence of maintenance teams on the factory floor but also to help factories reduce their inventory costs by alerting operators when machinery starts functioning outside nominal parameters, endangering the quality of the job in progress as well as the integrity of the machinery.
Built for Edge Impulse by the Zalmotek team