Flat Tire Example init: {tire(flat),tire(spare),at(flat,axle),at(spare,trunk)} goal: at(spare,axle) operators: –Remove(obj,loc) precond: at(obj,loc) effect:

Slides:



Advertisements
Similar presentations
Practical Planning: Scheduling and Hierarchical Task Networks Chapter CS 63 Adapted from slides by Tim Finin and Marie desJardins.
Advertisements

CSE391 – 2005 NLP 1 Planning The Planning problem Planning with State-space search.
Planning
Planning Module THREE: Planning, Production Systems,Expert Systems, Uncertainty Dr M M Awais.
Planning Module THREE: Planning, Production Systems,Expert Systems, Uncertainty Dr M M Awais.
Dana Nau: Lecture slides for Automated Planning Licensed under the Creative Commons Attribution-NonCommercial-ShareAlike License:
Classical Planning via Plan-space search COMP3431 Malcolm Ryan.
Problem Solving by Searching Copyright, 1996 © Dale Carnegie & Associates, Inc. Chapter 3 Spring 2007.
Plan Generation & Causal-Link Planning 1 José Luis Ambite.
Motion Planning for Point Robots CS 659 Kris Hauser.
Planning with Constraints Graphplan & SATplan Yongmei Shi.
Planning Graphs * Based on slides by Alan Fern, Berthe Choueiry and Sungwook Yoon.
TWEAK solving the Sussman anomaly ON(A,B) ON(B,C) ON(C,A) ONTABLE(A) ONTABLE(B) CLEAR(C) CLEAR(B) ARMEMPTY STACK(A,B) STACK(B,C) Two step additions to.
Planning Planning is fundamental to “intelligent” behaviour. E.g.
Sussman anomaly - analysis The start state is given by: ON(C, A) ONTABLE(A) ONTABLE(B) ARMEMPTY The goal by: ON(A,B) ON(B,C) This immediately leads to.
Notes 10: Planning; STRIPs Planning Systems ICS 271 Fall 2008.
Ch. 10 – Planning Supplemental slides for CSE 327 Prof. Jeff Heflin.
1 Classical STRIPS Planning Alan Fern * * Based in part on slides by Daniel Weld.
14 Jan 2004CS Blind Search1 Solving problems by searching Chapter 3.
1 Solving problems by searching Chapter 3. 2 Why Search? To achieve goals or to maximize our utility we need to predict what the result of our actions.
EXAMPLE OF PARTIAL ORDER PLANNING. Action Recipes: Action: Stack(x,y) Prec: Clear(x) ^ Clear(y) ^ OnTable(x) Effect: ~Clear(y) ^ ~OnTable(x) ^ On(x,y)
Nonholonomic Multibody Mobile Robots: Controllability and Motion Planning in the Presence of Obstacles (1991) Jerome Barraquand Jean-Claude Latombe.
CPSC 322 Introduction to Artificial Intelligence November 19, 2004.
Planning Copyright, 1996 © Dale Carnegie & Associates, Inc. Chapter 11.
Planning: Part 2 Partial Order Planning COMP151 April 2, 2007.
1 Solving problems by searching Chapter 3. 2 Why Search? To achieve goals or to maximize our utility we need to predict what the result of our actions.
Problem Solving by Searching Copyright, 1996 © Dale Carnegie & Associates, Inc. Chapter 3 Spring 2004.
Planning: Part 3 Planning Graphs COMP151 April 4, 2007.
GraphPlan ● (you don't have to know anything about this, but I put it here to point out that there are many clever approaches to try to find plans more.
1 Planning. R. Dearden 2007/8 Exam Format  4 questions You must do all questions There is choice within some of the questions  Learning Outcomes: 1.Explain.
1 Lecture 12 example (from slides prepared by Prof. J. Rosenchein)
Planning II CSE 473. © Daniel S. Weld 2 Logistics Tournament! PS3 – later today Non programming exercises Programming component: (mini project) SPAM detection.
Planning: Part 1 Representation and State-space Search COMP151 March 30, 2007.
Planning Copyright, 1996 © Dale Carnegie & Associates, Inc. Chapter 11.
Automated Planning and HTNs Planning – A brief intro Planning – A brief intro Classical Planning – The STRIPS Language Classical Planning – The STRIPS.
Planning in the Real World Time and Resources Hierarchical Task Network Conditional Planning Execution Monitoring and Replanning Continuous Planning MultiAgent.
1 Planning Chapters 11 and 12 Thanks: Professor Dan Weld, University of Washington.
1 Solving problems by searching Chapter 3. 2 Why Search? To achieve goals or to maximize our utility we need to predict what the result of our actions.
AI Principles, Lecture on Planning Planning Jeremy Wyatt.
Planning II CSE 573. © Daniel S. Weld 2 Logistics Reading for Wed Ch 18 thru 18.3 Office Hours No Office Hour Today.
PLANNING Partial order regression planning Temporal representation 1 Deductive planning in Logic Temporal representation 2.
22/11/04 AIPP Lecture 16: More Planning and Operators1 More Planning Artificial Intelligence Programming in Prolog.
1 Solving problems by searching This Lecture Chapters 3.1 to 3.4 Next Lecture Chapter 3.5 to 3.7 (Please read lecture topic material before and after each.
Planning (Chapter 10)
Computing & Information Sciences Kansas State University Lecture 22 of 42 CIS 530 / 730 Artificial Intelligence Lecture 22 of 42 Planning: Sensorless and.
Haley: A Hierarchical Framework for Logical Composition of Web Services Haibo Zhao, Prashant Doshi LSDIS Lab, Dept. of Computer Science, University of.
Planning (Chapter 10)
COMP322/S2000/L281 Task Planning Three types of planning: l Gross Motion Planning concerns objects being moved from point A to point B without problems,
AI Lecture 17 Planning Noémie Elhadad (substituting for Prof. McKeown)
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.
Notes 9: Planning; Strips Planning Systems ICS 270a Spring 2003.
1 CMSC 471 Fall 2004 Class #21 – Thursday, November 11.
1 CMSC 471 Fall 2002 Class #24 – Wednesday, November 20.
Planning I: Total Order Planners Sections
April 3, 2006AI: Chapter 11: Planning1 Artificial Intelligence Chapter 11: Planning.
Planning in FOL Systems sequences of actions to achieve goals.
Lecture 2: Problem Solving using State Space Representation CS 271: Fall, 2008.
Dana Nau: Lecture slides for Automated Planning Licensed under the Creative Commons Attribution-NonCommercial-ShareAlike License:
Planning as Satisfiability
Planning as Search State Space Plan Space Algorihtm Progression
Planning (Chapter 10) Slides by Svetlana Lazebnik, 9/2016 with modifications by Mark Hasegawa-Johnson, 9/2017
Planning (Chapter 10)
Introduction Contents Sungwook Yoon, Postdoctoral Research Associate
Planning (Chapter 10)
Problem Solving by Searching
Planning: Heuristics and CSP Planning
AI Planning.
Class #20 – Wednesday, November 5
Supplemental slides for CSE 327 Prof. Jeff Heflin
Prof. Pushpak Bhattacharyya, IIT Bombay
Presentation transcript:

Flat Tire Example init: {tire(flat),tire(spare),at(flat,axle),at(spare,trunk)} goal: at(spare,axle) operators: –Remove(obj,loc) precond: at(obj,loc) effect:  at(obj,loc),at(obj,ground) –PutOn(t,axle) precond: tire(t),at(t,ground),  at(flat,axle) effect:  at(t,ground), at(flat,axle) –LeaveOvernight precond: effect:  at(spare,ground),  at(spare,trunk),  at(spare,axle)  at(flat,ground),  at(flat,trunk),  at(flat,axle)

SatPlan propositionalization –create separate literals for each ground instance of each fluent for each state index e.g. on_a_b_1, on_a_b_2, on_a_b_3, gripper_empty_1, gripper_empty_2... –create propositions for action taken at step t: pickup_a_b_1... convert axioms in FOL to propositional sentences (one copy for each time index) –possibility axioms:  s,a Prec(s)  Poss(a,s) gripper_empty_1  clear_a_1  poss_pickup_a_1 gripper_empty_2  clear_a_2  poss_pickup_a_2 gripper_empty_1  clear_b_1  poss_pickup_b_1 –successor axioms:  s Poss(a,s)  Eff(result(a,s)) poss_pickup_a_1  pickup_a_1  holding_a_2  gripper_empty_2 –frame axioms:  s Flu(s)  a  CancelingActions  Flu(result(a,s)) on_b_c_1  poss_pickup_a_1  on_b_c_2 use a satisfiability solved like DPLL or WalkSat –query: poss_on_a_b_1? poss_on_a_b_2? poss_on_a_b_3? –the truth values of action propositions in the model tell you the plan pickup_c_a_1=T, puton_c_table_2=T, pickup_b_table_3=T, puton_b_a_4=T

more complex planning domains

Plan Abstraction –the idea: simplify problem by temporarily “dropping” easy preconditions defer solving them till later make high-level plan, then fill in details –system: ABSTRIPS –example: make a high-level plan for solving 8-puzzle where you assume you can move pieces without empty space

Hierarchical Task Networks (HTNs) –plan library a list of standard operating procedures, written out in procedural syntax ExecuteMissionWithHelicopters: –preconds: haveHelicopters, knowTargetCoordinates –steps: {FlyFlightPlan,Engage,ReturnToBase} –high-level operators vs. low-level operators –expansion, multiple choices of how to achieve subtask –still a problem of sub-goal interactions –system: STEAM

plan monitoring and repair –what if action in step i might fail? –non-deterministic domains (most in real-world) –at each step, check pre-conditions –if expected conditions not satisfied, re-invoke planner costly to re-plan from scratch search for modification of existing plan, e.g. re-do previous action or insert steps to get back on track contingent planning: –a plan with a branch in it –include sensor actions to sense state –algorithm to create such plans to achieve goals goal regression? universal planners: generate state-action tables, finite-state machines {pickupA, senseHolding, while not holding do {pickupA,senseHolding}} {swingAx, senseTree, while TreeStanding do {swingAx,senseTree}} enter room sense light off(light)?  flip(switch) next action

plan optimization –use heuristics to search for a plan that minimizes cost of operators, or time... –start by finding any plan that achieves goals, and then apply incremental modifications related to scheduling –critical path method (CPM) – compute [ES,LS] –resource constraints (e.g. 1 machine to add engine)

Robotics basic approach: –navigation in configuration space – find path from initial configuration to goal configuration –# dimensions = degrees of freedom (parameters that can be controlled, e.g. joint angles) –find path; avoid collisions with “obstacles”