We investigate small energy releases known as acoustic emissions (AEs) occurring during fault movements, akin to sounds produced in earthquakes. In laboratory settings, we gently shear a layer of silica powder to simulate crushed rock near the Earth's surface. Using a specially designed apparatus, we control stress, temperature, and humidity while recording AEs with external sensors. Calibration with simple ball-drop tests converts sensor signals into physical measurements. We observe that AEs persist even at very slow slip rates, indicating they likely reflect internal grain motion. This suggests a connection between AE intensity and changes in friction. Our findings offer new insights into fault behavior under low-stress conditions and lay the groundwork for using machine learning to analyze AE signals and predict frictional characteristics.