Course Description

Robots are increasingly integral to the contemporary economy. Specialized robots have notably enhanced efficiency, safety on the job, and the quality of products. Yet, these robots are generally designed for specialized tasks in controlled settings, lacking the ability to handle diverse tasks in real-world situations. How can we transition robots from limited-use scenarios to everyday life, where they can aid us in various tasks as helpers and companions? Achieving this will require a new form of general-purpose robot autonomy, where robots can interpret their surroundings through their own sensors and make well-informed decisions as a result. This course delves into cutting-edge machine learning and AI algorithms for autonomous robots. It explores advanced subjects focused on 1) how robots interpret unstructured environments using raw sensory input, 2) how robots make choices based on this understanding, and 3) how robots can continuously learn and adapt in the physical world.

Course Time and Location

Lecture: 10:10 AM-12:00 PM, Fridays
Location: 451 Mudd

Online Platforms

Canvas
Ed Discussion
Paper Presentation Scheduling

Learning Objective

This course is designed for graduate students who have a keen interest in robotics and AI. It is particularly relevant for those aiming to pursue research in this field. Through the curriculum, students will:

• Gain insights into the possibilities and social impacts of general-purpose robot autonomy in real-world settings, comprehend the technical obstacles involved in its development, and understand how machine learning and AI can help overcome these challenges.
• Become acquainted with a variety of model-based and data-driven approaches related to robot perception and decision-making.
• Develop the skills to assess, articulate, and implement sophisticated AI-driven methodologies to tackle issues in the realm of robotics.

Prerequisites

Students are expected to have the following backgrounds:

• A foundational understanding of data structures and algorithms, along with hands-on programming experience. Expertise in Python is mandatory, and a good understanding of C/C++ is advantageous.
• Being familiar with calculus, statistics, and linear algebra is essential, along with strong mathematical skills.
• Previous coursework or comparable experience in AI and Machine Learning is highly recommended.
• Be enthusiastic, patient, and courageous about working with robotics and AI systems.