A Microcontroller-Based System for Tool Condition Monitoring: Assessing the Viability of Low-Cost Vibration and Current Sensors
Gustavo Rogério Da Silva Barbosa, Leandro Pontes De Andrade, R.A. de Lima, João Fausto Lorenzato de Oliveira, Rogério Pontes de Araújo · IEEE Access · 2026
Tool condition monitoring (TCM) is essential for maintaining machining efficiency and workpiece quality. However, the prohibitive cost of industrial-grade instrumentation remains a barrier for small-to-medium enterprises. In this context, we evaluated a low-cost (approximately US ${\$}$ 30) multisensor architecture for in-process flank wear monitoring, assessing the individual diagnostic capabilities of vibration and electrical current sensing. The system integrates an MPU-9250 microelectromechanical systems (MEMS) accelerometer and an SCT013-000 current sensor with an ESP32 microcontroller, operating at a 2 kHz sampling rate via a wired UART serial interface. Vibration RMS exhibited a strong positive correlation with flank wear ( $\rho =0.94$ ), while frequency-domain analysis revealed statistically significant spectral centroid migration and harmonic power shifts consistent with tool degradation. Conversely, electrical current sensing yielded a negligible correlation ( $\rho = -0.09$ ), indicating that wear-induced load variations were overshadowed by the spindle’s baseline power demand under the tested cutting conditions. These findings suggest that the viability of low-cost current sensing depends on the machining regime, particularly the relative contribution of cutting power to total spindle load. Benchmarking confirmed that the proposed system achieves the highest sampling rate among reviewed MEMS-based academic prototypes while operating at approximately 15% of the cost of the closest comparable setup. Although the limited experimental dataset (n = 6) warrants further validation, these results indicate that vibration-based MEMS sensing is a promising low-cost modality for TCM in finishing operations, and that the diagnostic sensitivity of current sensing is contingent on the operating conditions.