Human Robot Interaction

Fall 2025. 16-867. Tuesday / Thursday 11:00am-12:20pm.

Announcements

Week 2

Sep 2 · 0 min read

Lecture on solving MDPs posted as well as a test-your-knowledge MDP example! Try to compute the values and policies for the simple MDP. Lecture notes on POMDPs also posted.

Course Overview

Robot interaction with people is inevitable: human engineers iteratively tune robot policies, autonomous cars navigate through crowded cities, construction workers teleoperate drones for building inspections, and assistive robots help end-users with daily living tasks.

In this graduate class, we will formalize such human-robot interaction (HRI) problems algorithmically. We will build the mathematical foundations for modeling human-robot interaction across robots and tasks, enable robots to understand human intent and predict human behavior, and study how robot learning changes in the presence of human feedback. The approaches covered will draw upon a variety of disciplines and tools such as sequential decision-making, cognitive science, Bayesian inference, and modern machine learning. Throughout the class, there will also be several guest lectures from experts in the field. Students will practice essential research skills including reviewing papers, writing project proposals, and technical communication.

In summary, in this class you will learn how to:

👩‍🔬 Mathematically model HRI
🧠 Predict human behavior & infer intent
🤝 Robot learning from human feedback
🙌 and more!

Prerequisites

The course is open to graduate students and advanced undergraduates. There are no official prerequisites but we expect familiarity with robotics and related AI topics (e.g., sequential decision-making / planning, basic machine learning); willingness to conduct a significant final project outside of class time; willingness to read scientific papers and engage in in-class discussions; willingness to conduct in-class presentations about scientific literature. Experience with high-level programming languages like Python are also strongly encouraged.

Schedule (Tentative)

Foundations

Aug. 26

Course Overview [Slides]: Syllabus

Aug. 28

Single-Agent Decision-Making [Notes]

Sept. 2

Value Iteration, Reinforcement Learning [Notes]: Test Your Knowledge (from CS188, Berkeley)

Sept. 4

POMDPs [Notes]

Sept. 9

Probability, Bayesian Inference: Goal Inference as Inverse Planning

Sept. 11

Intent Inference & Expression: Expressing Thought, Functional Expressive Motion

Sept. 16

Reward and Policy Learning

Sept. 18

Experimental Design & Statistical Analysis: HW #1 DUE A Primer for Conducting Experiments in Human–Robot Interaction

Prediction for Action

Sept. 23

Guest Lecture Trajectory Forecasting (Ingrid Navarro, CMU)

Sept. 25

Trajectory Forecasting: Paper Reading Confidence-Aware Prediction, ManiCast

Sept. 30

Collaboration, Assistance, & Coordination: Project Proposal Due

Oct. 2

Shared Autonomy: Paper Reading Shared Autonomy via Hindsight Optimization, LILA

Oct. 7

HRI as a Game

Oct. 9

HRI as a Game: Paper Reading Planning for AVs that Effect Humans, Long-term Robot Influence on Humans

Oct. 14

No Class Fall Break 🍂

Oct. 16

No Class Fall Break 🍂

Learning, Alignment, and Safety

Oct. 21

Representation Learning: Paper Reading RAPL, ALGAE

Oct. 23

Project Pitch & Red-Teaming: Mid-term Report Due Mid-Term Presentation Due

Oct. 28

Alignment

Oct. 30

Alignment: Paper Reading Deep RL from Preferences, Learning from Physical HRI

Nov. 4

No Class Democracy Day

Nov. 6

Guest Lecture HRI in the Era of Foundation Models (Sidd Karamcheti, TRI / Georgia Tech)

Nov. 11

Guest Lecture Active Learning (Erdem Bıyık, USC)

Nov. 13

Active Learning: HW #2 Due Paper Reading Active Learning from Critiques, Asking Easy Questions

Nov. 18

Safety & Uncertainty in HRI: Safety with Agency

Nov. 20

Safety & Uncertainty in HRI: Paper Reading Robots that Suggest Safe Alternatives, Robots that Ask for Help

Nov. 25

Cancelled Class Thanksgiving

Nov. 27

No Class Thanksgiving

Project Presentations

Dec. 2

Project Presentations: Slides Due 11:59 pm ET, Dec 1

Dec. 4

Project Presentations: Project Report Due Dec 11