The vibration-assisted atomic force microscope (AFM)-based nanomachining offers a promising opportunity for low-cost nanofabrication with high tunability. However, critical challenges reside in advancing the throughput and the quality assurance of the process due to extensive offline experimental investigations and characterizations, which in turn hinders the wide industry applications of current AFM-based nanomachining process. Hence, it is necessary to create an in-process monitoring for the nanomachining to allow real-time inspection and process characterizations. This paper reports a sensor-based analytic approach to allow real-time estimations of the AFM-based nanomachining process. The temporal-spectral features of collected acoustic emission (AE) sensor signals are applied to predict the depth morphology of nanomachined trenches under different machining conditions. The experimental case study suggests that the most significant frequency domain information from AE sensor can accurately predict (R-squared value around 92%) the nanomachined depth profile. It, therefore, breaks the current limitation of characterization tools at the nanoscale precision level, and opens up an opportunity to allow real-time estimation for quality inspection of vibration-assisted AFM-based nanofabrication process.