The Grainger College of Engineering
Electrical & Computer Engineering

ECE 484

ECE 484 - Principles of Safe Autonomy

Spring 2026

TitleRubricSectionCRNTypeHoursTimesDaysLocationInstructor
Principles of Safe AutonomyECE484AB176103LAB00900 - 0950 F  5072 Electrical & Computer Eng Bldg Sayan Mitra
Principles of Safe AutonomyECE484AB276104LAB01000 - 1050 F  5072 Electrical & Computer Eng Bldg Sayan Mitra
Principles of Safe AutonomyECE484AB376105LAB01100 - 1150 F  5072 Electrical & Computer Eng Bldg Sayan Mitra
Principles of Safe AutonomyECE484AB476106LAB01200 - 1250 F  5072 Electrical & Computer Eng Bldg Sayan Mitra
Principles of Safe AutonomyECE484AB576107LAB01300 - 1350 F  5072 Electrical & Computer Eng Bldg Sayan Mitra
Principles of Safe AutonomyECE484AB677479LAB01600 - 1650 F  5072 Electrical & Computer Eng Bldg Sayan Mitra
Principles of Safe AutonomyECE484AL173236LEC41100 - 1220 T R  1015 Electrical & Computer Eng Bldg Sayan Mitra

See full schedule from Course Explorer

Official Description

Introduces techniques for building autonomous systems such as autonomous cars, delivery drones, and manufacturing robots, and techniques for performing their safety analysis. Covers key algorithms and approaches in perception, modeling, motion planning, control, and safety analysis, with a view towards understanding their basic assumptions and performance guarantees. Also provides exposure to some of the state-of-the-art software tools for control, simulation, and analysis. Students will get experience through labs, programming assignments, and they will perform hands-on laboratory work on the Polaris GEM autonomous vehicle platform. Course material is distilled from recent research papers; thus, there is no required textbook. Course Information: 4 undergraduate hours. 4 graduate hours. Prerequisite: CS 124, ECE 220 or equivalent; ECE313, IE300, or STAT400. A course on data structures, algorithms, differential equations, and linear algebra is recommended.

Subject Area

  • Robotics, Vision, and Artificial Intelligence
  • AI: Robotics, Autonomous Systems, and AI

Course Director

Sayan Mitra

Description

Teaches seniors and first year graduate students in Electrical and Computer Engineering advanced topics in semiconductor device processing. Covers the principles of advanced methods of pattern delineation, pattern transfer, and modern material growth, and how these are applied to produce novel and high-performance devices and circuits in various electronic materials with special emphasis on semiconductors. Issues in computer simulation of processes and the manufacturing of devices and circuits are also covered.

Goals

By the end of this course, you will gain an understanding of the key concepts involved in designing and evaluating autonomous systems. These concepts can be organized around three core tasks: (a) sensing, perception, and state estimation; (b) decision making, planning, and control; and (c) evaluation and correctness arguments based on formal verification. Machine learning techniques play an important role across all three tasks. You will also gain hands-on experience implementing algorithms and models using state-of-the-art software tools. Finally, the course provides practical experience in engineering a complete autonomy stack for a sensor-rich platform—such as a full-sized vehicle, a quadrotor, or a scaled vehicle—with particular emphasis on rigorous evaluation.

Topics

Course topics (Spring 2026):

  • Overview of autonomous systems and an introduction to safety

  • Foundations of safety verification

    • verification concepts and specifications (requirements)

    • finite-state automata models

    • counterexamples and debugging safety failures

  • Reachability analysis and inductive invariants for safety

  • Perception fundamentals (machine learning + vision geometry)

    • neural networks and gradient-based learning

    • camera models and coordinate representations (intrinsic/extrinsic matrices, homogeneous coordinates)

    • calibration, projection, and linear-algebraic problems (eigenvalue formulation)

    • depth estimation and visual odometry

    • multi-view geometry (fundamental matrix, epipolar geometry)

  • Control for autonomy

    • ODE models, Lipschitz continuity, and bang-bang control

    • PID control

    • linear systems, stability, Lyapunov methods, Hurwitz criteria

  • Probabilistic state estimation and localization (filtering)

    • Markov chains, conditional probability, and motion/measurement models

    • Bayes filter, histogram filter, belief representations

    • Kalman filter

    • particle filter, importance sampling

    • SLAM overview and synthesis of filtering + mapping

  • Planning and search

    • graph search and uniform-cost search

    • A* and heuristic search (cost-to-go heuristics)

    • hybrid A*, PRM, and probabilistic completeness

    • sampling-based planning: RRT, RRG, asymptotic optimality

  • System platforms and applied autonomy contexts (GEM vehicle, F1tenth, GRAIC, drones)

  • Course integration and assessment milestones

    • review sessions and midterms

    • project review and guest lecture

    • final presentation/demo during finals period


Computer Usage

Workstations in ECE5072 have been setup with robotics specific tools and simulators. Students are expected to work on these workstations for their assignements.

Texts

S. Campbell, The Science and Engineering of Microelectronic Fabrication, 2nd ed., Oxford Press, 2001; chapters from several other books; numerous journal articles; proceedings of conferences; and industry reports

Instructional Objectives

You are expected to (a) attend lectures (TTh 11–12:30) and keep up with the material through the suggested exercises; (b) complete four homework and programming assignments; (c) attend Friday lab sections to discuss and work through assignments; (d) prepare for and take three in-class midterms; and (e) complete a substantial autonomy project. Additional details are provided below.

Last updated

1/20/2026