350 likes | 581 Views
16.412J/6.835 Intelligent Embedded Systems. Prof. Brian Williams Rm 37-381 Rm NE43-838 Williams@mit.edu. MW 11-12:30, Rm 33-418. Outline. Course Objectives and Assignments Types of Reasoning Kinds of Intelligent Embedded Systems A Case Study: Space Explorers. Plan. Monitor & Diagnosis.
E N D
16.412J/6.835 Intelligent Embedded Systems Prof. Brian Williams Rm 37-381 Rm NE43-838 Williams@mit.edu MW 11-12:30, Rm 33-418
Outline • Course Objectives and Assignments • Types of Reasoning • Kinds of Intelligent Embedded Systems • A Case Study: Space Explorers
Plan Monitor & Diagnosis Execute Course Objective 1 • To understand fundamental methods for creating the major components of intelligent embedded systems. Accomplished by: • First ten lectures on basic methods • ~ 5 problem sets during the first ten lectures to exercise basic understanding of methods.
Basic Method Lectures • Decision Theoretic Planning • Reinforcement Learning • Partial Order Planning • Conditional Planning and Plan Execution • Propositional Logic and Inference • Model-based Diagnosis • Temporal Planning and Execution • Bayesian Inference and Learning More Advanced: • Graph-based and Model-based Planning • Combining Hidden Markov Models and Symbolic Reasoning
Course Objective 2 • To dive into the recent literature, and collectively synthesize, clearly explain and evaluate the state of the art in intelligent embedded systems. Accomplished by: • Weekly thought questions (~ 2 page answers) • Group lecture on advance topic • 45 minute lecture • Short tutorial article on method 1-3 methods • Demo of example reasoning algorithm • Groups of size ~3.
Plan Monitor & Diagnosis Execute Course Objective 3 • To apply one or more reasoning elements to create a simple agent that is driven by Goals or Rewards Accomplished by: • Final project during last third of course • Implement and demonstrate one or more reasoning methods on a simple embedded system. • Short final presentation on project. • Final project report.
Outline • Course Objectives and Assignments • Types of Reasoning(Slides compliments of Prof Malik, Berkeley) • Kinds of Intelligent Embedded Systems • A Case Study: Space Explorers
Agents and Intelligence Prof Malik, Berkeley
Reflex agents Compliments of Prof Malik, Berkeley
Reflex agent with state Compliments of Prof Malik, Berkeley
Goal-oriented agent Compliments of Prof Malik, Berkeley
Utility-based agent Compliments of Prof Malik, Berkeley
Outline • Course Objectives and Assignments • Types of Reasoning • Kinds of Intelligent Embedded Systems • A Case Study: Space Explorers
Immobile Robots: Intelligent Offices and Ubiquitous Computing
courtesy NASA The MIR Failure
Portable Satellite Assistant courtesy NASA Ames
MIT Spheres courtesy Prof. Dave Miller, MIT Space Systems Laboratory
courtesy JPL Distributed Spacecraft Interferometers to search for Earth-like Planets Around Other Stars
A Goldin Era of Robotic Space Exploration courtesy JPL ``Our vision in NASA is to open the Space Frontier . . . We must establish a virtual presence, in space, on planets, in aircraft and spacecraft.’’ - Daniel S. Goldin, NASA Administrator, May 29, 1996
Model-based Embedded and Robotic Systems Group, MIT Cooperative Exploration Distributed Planning Group, JPL
MIT Model Based Embedded and Robotics GroupAutonomous Vehicles Testbed
Robotic Vehicles • ATRV Rovers • Monster Trucks • Blimps • Spheres • Simulated Air/Space Vehicles
Indoor test range • Aim & Scope: • indoor experiments for target site exploration • cooperative exploration
exploration feature path planned/taken way point Scenario Cooperative Target Site Exploration: Heterogeneous rover team and blimps explore science sites determined by remote sensing • Tasks: • small scout rovers (ATRV Jr) explore terrain as described in earlier scenarios • blimps provide additional fine grain air surveillance • scout rovers identify features for further investigation by sample rover (ATRV) • scout rovers provide refined terrain mapping for path planning of the larger sample rover • Scenario Research Objective • Extend coordination to heterogeneous team … exploration region identified feature goal position
Exploring life under Europa Cryobot & Hydrobot courtesy JPL
Outline • Course Objectives and Assignments • Types of Reasoning • Kinds of Intelligent Embedded Systems • A Case Study: Space Explorers
A Capable Robotic Explorer: Cassini Faster, Better, Cheaper • 150 million $ • 2 year build • 0 ground ops • 7 year cruise • ~ 150 - 300 ground operators • ~ 1 billion $ • 7 years to build Cassini Maps Titan courtesy JPL
courtesy JPL ``Our vision in NASA is to open the Space Frontier . . . We must establish a virtual presence, in space, on planets, in aircraft and spacecraft.’’ - Daniel S. Goldin, NASA Administrator, May 29, 1996
Four launches in 7 months Mars Climate Orbiter: 12/11/98 Mars Polar Lander: 1/3/99 QuickSCAT: 6/19/98 Stardust: 2/7/99 courtesy of JPL
Vanished: • Mars Polar Lander • Mars Observer courtesy of JPL Spacecraft require commonsense…
Quintuple fault occurs (three shorts, tank-line and pressure jacket burst, panel flies off). Mattingly works in ground simulator to identify new sequence handling severe power limitations. Mattingly identifies novel reconfiguration, exploiting LEM batteries for power. Swaggert & Lovell work on Apollo 13 emergency rig lithium hydroxide unit. Houston, We have a problem ... courtesy of NASA