| Part A: Robots that Plan and Act in the World |
| 1 |
Introduction to Cognitive Robots
Remote Explorers and Human Interation Systems |
|
| A1: Robots that Deftly Navigate |
| 2 |
Planning Routes by Generating Maps
Configuration Spaces, Visibility Graphs, Voronai Diagrams, Potential Fields, and Cell Decomposition |
|
| 3 |
Randomized Path Planning
Kino-dynamic Planning, Planning with Moving Obstacles, Probabilistic Roadmaps (PRMs), Rapidly Exploring Random Trees (RRTs) |
Problem set 1 out |
| A2: Planning and Executing Complex Missions |
| 4 |
Path Planning in Unknown Environments: An Overview |
|
| 5 |
Incremental Path Planning
Single Source Shortest Path, D*, LRTA* |
|
| 6 |
Mission-level Task Planning
Partial Order Planning, Constraint-based Interval Planning, and Simple Temporal Networks (STNs) |
Problem set 1 due
Problem set 2 out |
| Part B: Robots that are State-Aware |
| 7 |
Foundations of Estimation
Bayes Filters, Kalman Filters, and HMMs |
|
| B1: Robots that Find Their Way in the World |
| 8 |
Determining Location Through Particle Filters
MCMC Methods, Rejection Sampling, Importance Sampling, Metropolis, Particle Filters for Localization |
|
| 9 |
Learning Maps
Scan-matching, ICP, SLAM using Kalman Filters, Topological Maps, Fast-Slam |
|
| B2: Robots that Deduce and Control Their Internal State |
| 10 |
Model-based Programming and Model-based Diagnosis
Model-based Diagnosis |
Problem set 2 due
Problem set 3 out |
| 11 |
Conflict-directed Diagnosis and Probabilistic Mode Estimation
Consistency-based Diagnosis |
|
| 12 |
Incremental Mode Estimation and Hybrid Systems
Incremental Logical Inference, Trajectory Tracking for Constraint-based, Gaussian Filtering for Hybrid HMMs (K-Best and Rao-Blackwell Particle Filtering) |
|
| 13 |
Optimal CSPs and Conflict-directed A*
Constraint Satisfaction Problems and Conflict-directed A* Search |
|
| 14 |
Context-based Vision
Bill Freeman Guest Lecture |
Problem set 3 due |
| Fast Planning |
| 15 |
Planning as Heuristic Forward Search
FF Planning
Student Advanced Lectures
LPG: Local Search for Planning Graphs (Seung Chung) |
Problem set 4 out |
| 16 |
Student Advanced Lectures (cont.)
Fast Solutions to Constraint Satisfaction Problems (Robert Effinger and Dan Lovell) |
|
| Cooperative Planning |
| 17 |
Student Advanced Lectures (cont.)
Distributed CSPs and Task Assignment (Thomas Leaute and Justin Werfel) |
|
| 18 |
Student Advanced Lectures (cont.)
Distributed Reinforcement Learning and MDPs (Lars Blackmore and Steve Block) |
|
| Vision-based Exploration |
| 19 |
Student Advanced Lectures (cont.)
Vision-based SLAM (Soren Riisgaard) |
Problem set 4 due |
| 20 |
Student Advanced Lectures (cont.)
Information Based Adaptive Robotic Exploration (Morten Rufus Blas) and Uncertainty and Visual Exploration Alexander Omelchenko) |
|
| Part C: Robots that Preplan for an Uncertain Future |
| 21 |
Reactive Planning in Large State Spaces Through Decomposition and Serialization
Student Advanced Lectures (cont.)
SIFT SLAM Vision Details (Vikash Mansinghka) |
|
| 22 |
The Linear Programming Approach to Approximate Dynamic Programming (Guest Lecturer: Daniela Pucci de Farias)
Markov Decision Processes, Approximate Dynamic Programming and Linear Programming, Performance and Error Analysis, and Constraint Sampling |
|
| 23 |
Partially Observable Markov Decision Processes
POMDPs, Policy Trees and Value Iteration |
|
| 24 |
Approximate Solutions to POMDPs
Heuristics, Coastal Navigation, and Real World Apps |
|
| 25 |
Dynamic Scheduling and Execution
Temporal Plan Execution, Dynamic Scheduling, and Simple Temporal Networks |
|
| 26 |
Project Demonstrations
10 Minute Student Presentations |
Final projects due by end of day |