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 translate into damage to the mechanical structure of the CNC machine.
To address this challenge we will be developing an audio-based predictive maintenance solution that will make use of a machine learning model trained to analyze sounds made by a hot wire cutter during normal operation to detect when the wire gets stuck in the workpiece and stop the task before damaging the mechanical assembly.
Predictive maintenance is a field of AI that is constantly evolving and offers many benefits to those who adopt it. Syntiant's sound-based predictive maintenance solution is one example of how this technology can be used to improve industrial equipment. By monitoring the sound made by industrial machinery, the audio-based predictive maintenance model can detect when the critical failure is going to happen and stop the task before damage is done. This type of solution has the potential to greatly reduce unplanned downtime and increase overall equipment effectiveness.
Built for Edge Impulse by the Zalmotek team