Piezoelectric acoustic nano and microelectromechanical systems (MEMS) are a cornerstone technology for sensing and RF spectral processing and are now advancing on three fronts: higher operating frequency, larger power handling, and AI-enabled design automation. This short course reviews thin film piezoelectric platforms, dispersion and waveguiding engineering, and microfabrication that extend operation from MHz ultrasonics into the millimeter wave regime. We will highlights recent advances in piezoelectric acoustic devices for applications in millimeter-wave (mmWave) signal processing, ultrasound transducers, power resonators, sensors, and emerging technologies. We begin by introducing transferred thin-film lithium niobate (LN) and sputtered scandium aluminum nitride (ScAlN) as platforms for mmWave acoustic resonators and filters, demonstrating record-high electromechanical coupling and quality factors at frequencies up to 60 GHz. We then examine high-power operation and reliability, focusing on acoustomigration, self-heating, spurious mode growth, and material and interface degradation, and we present design and test methods that increase power handling without sacrificing quality factor, bandwidth, and frequency stability. For AI design-measurement, we introduce data-driven models that map structure to key figures of merit and to the full admittance spectrum, enabling screening and optimization. We highlight sparse to dense reconstruction that recovers a full spectrum from as few as 16 evenly spaced sweep points, and inverse synthesis using a physics-guided surrogate library to propose new, spurious suppressed resonators. Finally, we discuss physics-guided simulation to real learning that corrects simulator bias under limited measurements.
