AA598 Decision-making and Control for Safe Interactive Autonomy

Instructor: Karen Leung
Time: Mon & Wed, 10:00 am - 11:20 am
Location: GUG 305
Office hours: Monday 12:00PM - 1:00PM, GUG 311B
Syllabus: Link Note: Subject to minor changes to adapt to the course progression

Overview

As we move towards a future where autonomy plays an integral role in our everyday lives, such as autonomous mobility and service robots, we must recognize that the success and impact of autonomy on society are intimately connected with the humans’ experience of interacting with the system. In particular, we must ensure the interactions between autonomous systems and humans are safe.

In this course, we will explore, investigate, and present a suite of methods and techniques for decision-making and control that contribute to safe autonomous operations, including developing models for human behaviors, reasoning about uncertainty, safe control, and safety analysis.

Prerequisites: Familiarity with linear algebra, differential equations, probability, feedback control, machine learning, deep learning, and programming (Python). AA/EE/ME 548 recommended.

For those registered, please fill out the following background knowledge survey to help us understand your familiarity with key topics essential for this course. While this course will cover some basic fundamentals, some level of familiarity in these topics is expected. Your responses will assist in calibrating the course content to better suit your needs.

Homework

There will be 3 homework. Submit on Canvas. Click here for more instructions.

Paper discussion and reviews

During the course you will need to:

  • (Group work) Lead a long paper discussion
  • Submit various short paper summaries
  • Submit talk summaries and reflections

Click here for more instructions.

Project

Click here for more instructions.

Guest lectures

Click here for more details about the guest lectures.

Useful references and resources

News

Jul 28, 2024

Course website launched!

Sep 25, 2024

Canvas now published. Homework 1 is published. Sign up to give a long paper discussion. First one is happening Oct 14th! Sign up sheet here!

Oct 23, 2024

Homework 2 is published. Please report any typos or issues. And more information about project proposal is on the Project page.

Oct 30, 2024

Some other links that was mentioned in class today (Oct 30). Game theory game show, Game-theoretic car racing, FaSTrack, Game theory demo.

Nov 07, 2024

Small update to hw2_starter_code_sqp.ipynb. You can pull the latest version, or update this line in the 5th cell objective += beta1 * cp.sum_squares(us) * markup**t with objective += beta1 * cp.sum_squares(us[t]) * markup**t

Nov 14, 2024

Homework 3 out. Talk details for guest lecture for safe controls module available.

Nov 19, 2024

More project details out. See project page for more details. Sign up for a presentation time slot here: Sign up sheet


Schedule

Subject to change

Week Date Topic Due Information
0 Wed, Sep 25 Introduction & Fundamentals I Background knowledge survey Lecture 1 slides
1 Mon, Sep 30 Fundamentals II   Lecture 2 slides
1 Wed, Oct 2 Fundamentals III   Lecture 2-3 notes
2 Mon, Oct 7 Human prediction models I   Lecture 4 slides
2 Wed, Oct 9 Human prediction models II   Lecture 5 slides
3 Mon, Oct 14 Long paper discussion (prediction) Short paper summaries [Paper 1] Kretzschmar et al 2016 Socially compliant navigation via IRL
[Paper 2] Rhinehart et al 2019 Goal-conditioned deep multi-agent trajectory forecasting (slides)
3 Wed, Oct 16 Human prediction models III   Lecture 6 slides
Quickfire summaries (Prediction)
3 Fri, Oct 18   Homework 1  
4 Mon, Oct 21 Interaction-aware planning I   Lecture 7 slides
4 Wed, Oct 23 Guest lecture: Dr. Boris Ivanovic Talk summary Talk details (Recording)
Lecture 7 slides cont’d
5 Mon, Oct 28 Interaction-aware planning II   Lecture 8 slides
5 Wed, Oct 30 Long paper discussion (planning) Short paper summaries [Paper 3] Fredovich-Keil, Bajscy et al 2020 Confidence-aware prediction & planning
[Paper 4] Burger et al 2022 Interaction-Aware Game-Theoretic Motion Planning
5 Fri, Nov 1   Project proposal  
6 Mon, Nov 4 Guest lecture: Prof. David Fridovich-Keil Talk summary Talk details (Recording)
Lecture 9 slides
6 Wed, Nov 6 Interaction-aware planning III   Lecture 10 slides
6 Fri, Nov 8   Homework 2  
7 Mon, Nov 11 (No class, Veteran’s Day)    
7 Wed, Nov 13 Safe control I   Lecture 11 slides
8 Mon, Nov 18 Long paper discussion (control) Short paper summaries [Paper 5] Tian et al 2021 Confidence-aware safety models
[Paper 6] Cosner et al 2022 Safety-Aware Preference-Based Learning for Safety-Critical Control
8 Wed, Nov 20 Guest lecture: Prof. Sylvia Herbert Talk summary  
9 Mon, Nov 25 Safe control II    
9 Wed, Nov 27 No lecture    
9 Fri, Nov 29   Homework 3  
10 Mon, Dec 2 Final presentations I    
10 Wed, Dec 4 Final presentations II    
Finals Wed, Dec 11   Final report  
Finals Fri, Dec 13   Homework & talk summaries Â