Abstract:
Translating theoretical kinematics into functional, moving hardware is one of the most exciting—yet challenging—hurdles in mechanical design. For early-career engineers and university students, the leap from mathematical models to physical prototypes often comes with a steep, frustrating learning curve. This presentation showcases how to bridge that gap by utilizing interactive visual synthesis tools, specifically MotionGen. By bringing abstract equations to life on screen, designers can rapidly iterate, optimize, and experiment with linkage behaviors long before committing to fabrication. Beyond the prototyping lab, this talk explores how these dynamic digital environments inject new energy into higher-education curricula, specifically within the *Mechanics of Machinery* course. Through a structured, project-based approach, the design process is accelerated, physical mechanical failures are minimized, and theoretical software simulations are confidently transformed into tangible engineering realities.

Abstract:
Many problems in engineering science share a difficult feature: the governing physics is only partially known, and experimental data is scarce or noisy. Classical simulation methods such as CFD are mature and accurate but require a complete description of the physics and are computationally demanding, while purely data-driven machine learning is data-hungry and yields non-physical predictions. Physics-Informed Neural Networks (PINNs) bridge this gap by embedding physical laws – expressed as partial differential equations – directly into the training of a neural network, making them well suited to the ubiquitous regime of limited physics and limited data.
This talk gives a self-contained introduction to PINNs, from the underlying principles to their practical implementation, and illustrates their power on two engineering applications: recovering three-dimensional gravity currents from sparse experimental data and identifying rheological parameters in granular flows. Together these examples show how PINNs offer a robust, GPU-efficient alternative to classical data assimilation.