<p>In the manufacturing sector, there is a growing demand for in-line process monitoring of machine tools. Scaling up this practice with machine learning enables tool life prediction and optimization, enhancing product quality while reducing manufacturing and maintenance costs. This paper investigates the practical application of cost-effective micro-electro-mechanical systems (MEMS) accelerometers for detecting tool wear in turning and milling processes. A classification approach using neural networks (NNs) is applied for tool wear detection, in which two distinct classes of tool wear sizes corresponding to specific thresholds of tool flank wear are defined. In this study, the assessment of the performance of MEMS accelerometers was primarily performed on the turning process. A novel sensorized tool holder was designed and developed for this purpose. This tool holder is capable of integrating multiple sensors, including accelerometers to enable comprehensive monitoring of the cutting process. This research involved selecting and rigorously assessing a series of MEMS accelerometers for tool wear detection, primarily in the turning process, using the tool holder as an experimental platform. The MEMS accelerometers exhibiting superior performance in the turning process were further evaluated for tool wear detection and process monitoring in milling operations. The results showed that MEMS accelerometers were capable of detecting tool wear in both milling and turning processes, while also effectively capturing additional parameters, such as bearing frequencies in milling with high accuracies.</p>