18-448 Computational Design of Cyber-Physical Systems

12 units, Carnegie Mellon University

Please refer to the ECE website for descriptions of each topic: https://courses.ece.cmu.edu/ Description for section A in Spring 2026: Most machines in the world around us, from vehicles to medical devices to robots, are designed manually. Yes, computer-Aided Design (CAD) software helps to digitize intent, but those tools only help improve the fidelity of engineers' direct specifications. As computational tools become more capable, it opens the possibility to the automated design of cyber-physical systems --- machines invented partially or entirely by AI. This course takes a pedagogical view of the computational design stack, with an emphasis on 'things that move.' Topics covered include: -Mathematical Preliminaries -Physical Simulation -Control Synthesis -Design Optimization -Machine Learning For Design -Designing For (Digital) Manufacturing -Interactive Workflows Each unit builds upon the last, culminating in a toolset that will empower students to design complex robots with just a couple dozen lines of code. The course includes a lab component and research-style final project.

Prerequisites: none listed

Offered: Spring 2027, Spring 2026, Fall 2025

Courseweave needs JavaScript to show sections, workloads and planning tools.