PLANNING IN AI. Determine the set of steps that are necessary to achieve a goal Some steps might be conditional, i.e., they are only taken when a set.

Slides:



Advertisements
Similar presentations
Planning
Advertisements

Artificial Intelligence 2005/06 Partial Order Planning.
1 Planning Chapter CMSC 471 Adapted from slides by Tim Finin and Marie desJardins. Some material adopted from notes by Andreas Geyer-Schulz,
Artificial Intelligence By Mr. Ejaz CIIT Sahiwal.
Artificial Intelligence Presentation
Planning  We have done a sort of planning already  Consider the “search” applied to general problem solving  The sequence of moves with the “Jugs” was.
Planning II: Partial Order Planning
CLASSICAL PLANNING What is planning ?  Planning is an AI approach to control  It is deliberation about actions  Key ideas  We have a model of the.
Plan Generation & Causal-Link Planning 1 José Luis Ambite.
Biointelligence Lab School of Computer Sci. & Eng. Seoul National University Artificial Intelligence Chapter 8 Uninformed Search.
Best-First Search: Agendas
Artificial Intelligence II S. Russell and P. Norvig Artificial Intelligence: A Modern Approach Chapter 11: Planning.
1 Classical STRIPS Planning Alan Fern * * Based in part on slides by Daniel Weld.
Problem Solving and Search in AI Part I Search and Intelligence Search is one of the most powerful approaches to problem solving in AI Search is a universal.
ARTIFICIAL INTELLIGENCE [INTELLIGENT AGENTS PARADIGM] Professor Janis Grundspenkis Riga Technical University Faculty of Computer Science and Information.
Artificial Intelligence 2005/06
Planning Copyright, 1996 © Dale Carnegie & Associates, Inc. Chapter 11.
1 Using Search in Problem Solving Part II. 2 Basic Concepts Basic concepts: Initial state Goal/Target state Intermediate states Path from the initial.
Artificial Intelligence Chapter 11: Planning
Uninformed Search Reading: Chapter 3 by today, Chapter by Wednesday, 9/12 Homework #2 will be given out on Wednesday DID YOU TURN IN YOUR SURVEY?
Planning II CSE 473. © Daniel S. Weld 2 Logistics Tournament! PS3 – later today Non programming exercises Programming component: (mini project) SPAM detection.
Planning Copyright, 1996 © Dale Carnegie & Associates, Inc. Chapter 11.
CSE (c) S. Tanimoto, 2008 State-Space Search 1 State-Space Search Motivation: Understanding “dynamics” in AI Statics versus Dynamics of AI Statics:
1 Planning Chapters 11 and 12 Thanks: Professor Dan Weld, University of Washington.
Distributed Scheduling. What is Distributed Scheduling? Scheduling: –A resource allocation problem –Often very complex set of constraints –Tied directly.
Planning II CSE 573. © Daniel S. Weld 2 Logistics Reading for Wed Ch 18 thru 18.3 Office Hours No Office Hour Today.
Artificial Intelligence Course outline Introduction Problem solving Generic algorithms Knowledge Representation and Reasoning Expert Systems Uncertainty.
Classical Planning Chapter 10.
Course Overview  What is AI?  What are the Major Challenges?  What are the Main Techniques?  Where are we failing, and why?  Step back and look at.
1 Plan-Space Planning Dr. Héctor Muñoz-Avila Sources: Ch. 5 Appendix A Slides from Dana Nau’s lecture.
Artificial Intelligence Lecture 9. Outline Search in State Space State Space Graphs Decision Trees Backtracking in Decision Trees.
Introduction to AI Robotics Chapter 2. The Hierarchical Paradigm Hyeokjae Kwon.
Homework 1 ( Written Portion )  Max : 75  Min : 38  Avg : 57.6  Median : 58 (77%)
Dana Nau: Lecture slides for Automated Planning Licensed under the Creative Commons Attribution-NonCommercial-ShareAlike License:
CS.462 Artificial Intelligence SOMCHAI THANGSATHITYANGKUL Lecture 07 : Planning.
Introduction to search Chapter 3. Why study search? §Search is a basis for all AI l search proposed as the basis of intelligence l all learning algorithms,
Midterm Review Prateek Tandon, John Dickerson. Basic Uninformed Search (Summary) b = branching factor d = depth of shallowest goal state m = depth of.
CPS 270: Artificial Intelligence Planning Instructor: Vincent Conitzer.
For Monday Read chapter 4, section 1 No homework..
Search CPSC 386 Artificial Intelligence Ellen Walker Hiram College.
1 Chapter 16 Planning Methods. 2 Chapter 16 Contents (1) l STRIPS l STRIPS Implementation l Partial Order Planning l The Principle of Least Commitment.
Introduction to Planning Dr. Shazzad Hosain Department of EECS North South Universtiy
MDPs (cont) & Reinforcement Learning
AI Lecture 17 Planning Noémie Elhadad (substituting for Prof. McKeown)
Introduction to Artificial Intelligence Class 1 Planning & Search Henry Kautz Winter 2007.
Problem Reduction So far we have considered search strategies for OR graph. In OR graph, several arcs indicate a variety of ways in which the original.
Intro to Planning Or, how to represent the planning problem in logic.
Searching for Solutions
Robust Planning using Constraint Satisfaction Techniques Daniel Buettner and Berthe Y. Choueiry Constraint Systems Laboratory Department of Computer Science.
Planning I: Total Order Planners Sections
Biointelligence Lab School of Computer Sci. & Eng. Seoul National University Artificial Intelligence Chapter 8 Uninformed Search.
Introduction to Artificial Intelligence
Planning: Heuristics and CSP Planning
Planning: Heuristics and CSP Planning
Class #17 – Thursday, October 27
CS344 : Introduction to Artificial Intelligence
Planning José Luis Ambite.
Graphplan/ SATPlan Chapter
CIS 488/588 Bruce R. Maxim UM-Dearborn
Class #19 – Monday, November 3
Chapter 6 Planning-Graph Techniques
Class #20 – Wednesday, November 5
Graphplan/ SATPlan Chapter
CS344 : Introduction to Artificial Intelligence
Graphplan/ SATPlan Chapter
Search.
Search.
An Introduction to Planning Graph
Prof. Pushpak Bhattacharyya, IIT Bombay
[* based in part on slides by Jim Blythe and Dan Weld]
Presentation transcript:

PLANNING IN AI

Determine the set of steps that are necessary to achieve a goal Some steps might be conditional, i.e., they are only taken when a set of conditions is present during plan execution.

TYPICAL PLANNING PROBLEMS * Representing and reasoning about time, causality, and intentions Physical and other kinds of constraints on acceptable solutions Uncertainty in the execution of plans How the “real world” is sensed and perceived Multiple agents who may cooperate or interfere * Lucci, Stephen; Kopec, Danny: Artificial Intelligence in the 21st Century

TERMINOLOGY Operator Schema The sequence of steps followed to solve a planning problem Characterize actions/events

STRIPS (STANFORD RESEARCH INSTITUE PROBLEM SOLVER) Operator Schema Consist of: Precondition Delete List Add List Variables in the Precondition, Delete List, and Add List that are bound at run-time Example: Pickup(X) Precondition: OnTable(X) ^ HandEmpty ^ Clear(X) Delete List: OnTable(X) HandEmpty Clear(X) Add List: Holding(X)

SOME TYPES OF PLANNING Hierarchical Planning Goals are not equally valued Some might be necessary Others might be desirable, but not necessary Numerical values provide hierarchical value Opportunistic Planning Exploit the conditions in a plan state to more easily achieve a goal Conditional Planning Planning based on things that “might happen.” Partial Order Planning Not all steps in the plan must be made in order

PLANNING STRATEGIES Means-Ends Analysis Reduce the distance between the current state and the goal state (GPS – Newell and Simon) Least Commitment Planner Only commit to a plan when forced by constraints Can maintain multiple/flexible “parallel” plans until forced to select a concrete alternative Ordering of steps is deferred as long as possible Depth-First Backtracking (“lifting”) Generates combinations of steps until one that works is found Impractical for large plans (exponential time)

PLANNING STRATEGIES (CONTINUED) Beam Search Expand the top few (beam width) nodes at each level of a breadth-first search One-The-Best Backtracking Rely on local information to determine a best-guess path, then backtrack – looking for good alternative candidates along the path Dependency-Directed Search Store dependencies between decisions, assumptions, and alternatives for each On failure, keep parts of the solution that are not dependent of the cause of failure

PLANNING STRATEGIES (CONTINUED) Opportunistic Search Favor the most constrained operations Meta-level Plans Plans about plans Techniques of planning can be selected dynamically based on the type of problem at hand. Distributed Planning Different parts (subplans) of the plan are developed by different “expert” components of the system

GRAPHPLAN Graph builds a graph starting from the initial state Uses “layered” or “parallel” plans, where each layer provides a set of actions that can be performed simultaneously. Mutually exclusive actions are identified at the level where they first occur. Builds a graph where each node represents the state after all possible transitions from the previous state. Solutions are found by working backward from the final node GraphPlan can also be used to generate heuristics for A* search notes/graphplan.pdf notes/graphplan.pdf