Spindle performance monitoring via accelerometer measurements in data-driven models

The future of manufacturing depends on transitioning traditional machines into intelligent machine tools that can monitor and control themselves. As the spindle is an essential component of machine tools, the performance of machine tool spindles should be tracked for quality control. For example, smart spindles could be equipped with accelerometers for monitoring the spindle performance via models that relate the measured accelerations to the spindle error motions. Various data-driven models were created that estimate spindle-related displacements from on-machine accelerations. The estimated displacements were compared, revealing the advantages and disadvantages of each model to monitor the spindle performance.