A machine wears along a noisy path toward a failure threshold, and its digital twin learns how fast from periodic inspections. Change the true wear rate, the fleet prior, the random fluctuation, the inspection noise and interval to watch the Kalman-updated drift estimate converge and the predicted remaining-useful-life density, an inverse-Gaussian mixture, sharpen around the true remaining life. A Blender-built bearing machine and probe, with numerical checks built in.