Classical Single and Multi-Agent Planning: An Introductory Tutorial. Ronen I. Brafman Computer Science Department Ben-Gurion University

Size: px
Start display at page:

Download "Classical Single and Multi-Agent Planning: An Introductory Tutorial. Ronen I. Brafman Computer Science Department Ben-Gurion University"

Transcription

1 Classical Single and Multi-Agent Planning: An Introductory Tutorial Ronen I. Brafman Computer Science Department Ben-Gurion University

2 Part I: Classical Single-Agent Planning

3 What Is AI Planning! A sub-field of AI concerned with modelling and answering the following problem: how to enable an artificial agent to autonomously choose good behaviours.! Original motivation: autonomous robots that achieve goals without being provided with an explicit program! R2D2, Get me some coffee! Instead of telling the agent what to do, we want to tell it what to achieve, and it will figure out what to do on its own.! Studies: models, languages, and algorithms for describing and solving different types of sequential decision problems! Many different applications in support of autonomy and task automation

4 Some AI Planning Applications

5 Focuses on Generic Methods (Domain Independence)! In principle, one could write special planners for different settings! Such planners can be very efficient, but:! A lot of effort to develop! Requires much expertise But! AI planning focuses on generic algorithms and language! Domain knowledge can be provided to planners via constraints that are also expressed in a general language 5

6 NASA s Experience Galileo's Jupiter or Cassini's Saturn missions:! $1G budget! Ground crew of Mars micro-rover Sojourne exploited AI/autonomy:! $100M budget! Much smaller ground team! Sojourne was designed to operate for two month, but lasted much longer Soujourne s level of autonomy is still low compared to what we can achieve!!

7 NASA s Experience Space exploration should be:! Low cost & rapid development: use off-the-shelf algorithms/ systems as much as possible! Low cost control: autonomous operations for long period of times! Smart algorithms can optimize success probability given tight deadlines and resource constraints Main Conclusion: low cost, off-the-shelf components for autonomous operation can lead to significant cost savings

8 AI Planning: Components Input:! World model:! Variables describing the state of the world! Initial/Current state of the world! Action model describes how the agent s ability to influence the world! Performance criteria: goal states, good states, bad states etc. Output:! Good behaviour

9 Rich Set of Models! Uncertainty! Stochastic actions! Partial observability and sensing! Temporal actions! Concurrency/interactions! Multi-agent! 9

10 Our Focus: Classical AI Planning! Single agent! Actions are deterministic and instantaneous! Initial state is fully known! The goal is to make some set propositions true! I.e. Reach some goal states! Behaviour = plan: a sequence of actions that, when applied in the initial state lead to one of the goal states

11 Classical Planning: An Informal Example Logistics Planning! World model:! Variables describing the location of trucks, planes, and packages! Initial state: current location of each object! Action model! Actions include load/unload packages and move vehicles! Model describes how each action impacts value of state variables! Goal: A state in which the packages are in the destination Output:! A sequences of actions move, load, unload, fly, etc. possibly concurrent, that cause the packages to end up in their destinations

12 Key Notions! Model: describes the relevant aspects of the world! Should be intuitive! May be very large and difficult to work with! Language: Tool for describing the world implicitly! Should be easy to describe intuitive/natural models! Interpretation: maps between language and model! Queries: questions about the model we wish to answer! Algorithms: compute answers to queries

13 Key Elements of Classical Planning! Model: Transition system! Set of states and transitions between them! Each transition has a label the action that brings it about! Each action has a cost (1 by default)! Initial state + Goal states! Language and Interpretation: We ll see! Query: Find a plan = sequence of actions/labels that takes us from the initial state to a goal state! Algorithms: path finding algorithms?

14 Example: Blocks World! States: possible configurations of 3 blocks on a table!actions: move a block from one block, or from the table, on top of another

15 Classical Planning Model = Automaton! States of the world = automaton state! Actions = letters/edge labels! Initial state = Initial state! Goal states = Absorbing state! Plan = a word in the language of the automata Simple solution: shortest path algorithm on automata graph

16 The Curse of Dimensionality Blocks States Even for a relatively small number of objects we cannot realistically construct, store, or compute with the explicit state transition graph

17 What Can We Do?! Obviously, we cannot construct the entire statetransition diagram, nor maintain it in memory, nor run algorithms on this explicit structure! We need to describing the model implicitly! We need techniques for generating path to the goal that explore only a small fragment of the model 17

18 Language: STRIPS [Fikes & Nilsson, 71]! A concise language for describing classical AI planning problems! States described via a set P of propositions! Any assignment to P is a possible world! Initial state: Simply one such assignment! Goal condition: partial assignment over P! Any state satisfying it is a goal state

19 STRIPS! Actions described by two lists! Preconditions: list of propositions! Must be true for the action to be applied in a state! Effects: list of literals! Become true following the action s application! Propositions not mentioned in the effects list remain unchanged

20 Example 1: BlocksWorld! State variables: on(r,b),on(b,r),on(r,g) on(g,r),on(g,b),on(b,g),on(r,table), on(g,table),on(b,table),clear(b),clear(r), clear(g),clear(table)! Actions: Move(X,Y,Z)! Preconditions: on(x,y),clear(x), Z!=Table, clear(z)! Effects: -on(x,y),on(x,z),-clear(z),clear(y)! Actions: Move(X,Y,Table)! Preconditions: on(x,y),clear(x)! Effects: -on(x,y),on(x,z),clear(y) 20

21 Example 2: The Logistics Domain

22 Example: The Logistics Domain! Every states of the world describes the location of packages, trucks and airplanes! Initial state:! package p1 is in the post-office in Chania, p2 is in Hotel Halepa, and p3 is in Beer-Sheva! Truck in Heraklion airport, and a truck in Beer-Sheva! Airplane in Heraklion airport! Goal: p1 and p2 in Beer-Sheva, p3 in Hotel Halepa! Actions: we can load and unload packages into/from a truck or a plane, we can move trucks between locations in the same country, and we can fly planes between airports

23 Logistics Domain Example! Propositions: at(p1,chaniapo), at(p1,halepa), at(p1,heraklionap), at(p1, TLV), at(p1, Beer-Sheva), in(p1,crete-truck), in(p1, isra-track), in(p1, plane), at(cretetruch, ChaniaPO),! Initial state (propositions assigned true): at(p1,chaniapo), at(p2,halepa), at(p3,bs), at(crete-truck, HeraklionAP), at(isra-truck, BS), at(plane, HeraklionAP)! Goal: at(p1,bs) & at(p2,bs) & at(p3, Halepa)

24 Logistics Domain Example Actions:! Load(package,vehicle,location)! Preconditions: at(package,location), at(vehicle,location)! Effects: -at(package, location), in(package, vehicle)! Unload -- similar! Move(vehicle,loc1,loc2)! Preconditions: at(vehicle, loc1)! Effects: -at(vehicle,loc1), at(vehicle, loc2)

25 Applying an Action! To apply an action at a state:! First, check if the preconditions hold! Next, update the value of the variables in the effect! All other variables state the same 25

26 Action Application: Example State: at(p1,chaniapo), at(p2,halepa), at(p3,bs), at(crete-truck, HeraklionAP), at(isra-truck, BS), at(plane, HeraklionAP)! Apply Load(p1,crete-truck,Halepa)! Preconditions: at(p1,halepa), at(crete-truck,halepa) not satisfied! Action is inapplicable 26

27 Action Application: Example State: at(p1,chaniapo), at(p2,halepa), at(p3,bs), at(crete-truck, HeraklionAP), at(isra-truck, BS), at(plane, HeraklionAP)! Apply Move(crete-truck,HeraklionAP,Halepa)! Preconditions: at(crete-truch,heraklionap) satisfied! Effect: at(crete-truck,halepa), -at(crete-truck,heraklionap)! New state: at(p1,chaniapo), at(p2,halepa), at(p3,bs), at(crete-truck, HeraklionAP), at(crete-truck, Halepa), at(isra-truck, BS), at(plane, HeraklionAP) 27

28 Planning Is Not Easy! STRIPS planning is hard:! PSPACE-Complete [Bylander, 1994]! NP-complete for practical purpose! Tractability?! (i.e., plans that are not too long)! In special cases with special structure under severe restrictions (e.g single effect + more restrictions)! Or under strong decomposability properties (2 nd part)

29 Planning in Practice! Many interesting and very nice algorithms! Partial order planning! Graphplan! Regression! Planning via Satisfiability! State-of-the-art: planning via heuristic forward search! Key: all planning algorithms require search! Best heuristics exist for forward search

30 Planning via Heuristic Forward Search! Start with the initial state! Expand the most promising state! Until a goal state is found! Most promising state: varies with the search algorithm, but generally speaking the state with the best heuristic value

31 Forward Search Space A B C {on&a&b,on&b&c,clear&a,on&table&c}" Move&to&table&a&b" Ini8al"State" {on&table&a,on&b&c,clear&a,clear&b,on&table&c}" Move&from&table&a&b" Move&to&table&b&c" Move&b&c&a" Goal"Condi8ons" {on&a&b,on&b&c,clear&a,on&table&c}" {on&table&a,on&table&b,"on&table&c,clear&a,clear&b,clear&c}" {on&table&a,on&b&a,on&table&c,clear&b,clear&c}" B A C 1"

32 Two Key Issues! How to compute the heuristic value! Key idea: solve a simple problem and use the solution to inform your choices in the original problem! We ll see two methods there are many others! How to use the heuristic value to improve search! There are many heuristic search algorithms with different properties I assume you know about search! A*, Best-first, hill-climbing, Weighted A*,

33 Heuristics! Heuristic functions should be fast to compute! Computed for each search state! Heuristic functions should be as accurate as possible! To find optimal plans we need admissible heuristics! Admissible: optimistic ; does not overestimate cost

34 Two Methods! Pattern Database Heuristic!! Landmark heuristic

35 Pattern Databases! Before search:! Project original problem to a subset of variables! Generate complete transition system! Compute shortest path from each state to the goal and store it! During search! With each state, associate the distance from its projection to the (projected) goal as its heuristic value

36 Pattern database heuristics ALR ARL LLR RRL ALL ARR LRR LLL RRR RLL BLL BRR LRL RLR BRL BLR Logistics problem with one package, two trucks, two locations: state variable package: {L, R, A, B} state variable truck A: {L, R} state variable truck B: {L, R} 36

37 Pattern database heuristics Abstraction induced by π {package} : ALR ARL LLR RRL ALL ARR LRR LLL RRR RLL BLL BRR LRL RLR BRL BLR h {package} (LRR) = 2 37

38 Pattern database heuristics Abstraction induced by π {package,truck A} : ALR ARL LLR RRL ALL ARR LRR LLL RRR RLL BLL BRR LRL RLR BRL BLR h {package,truck A} (LRR) = 2 38

39 39 h {Package,Truck A} h {Package} State 2 2 LRR 2 2 LLR 2 2 LLL 2 2 LRL 2 1 ALR 2 1 ALL 1 1 ARL 1 1 ARR 1 1 BLR 1 1 BLL 1 1 BRL 1 1 BRR 0 0 RRR 0 0 RLR 0 0 RLL 0 0 RRL π { } π {, }

40 {Package,Truck A} Best first search with h LRR Iter. Fringe LLR LRL 0 1 LRR(2) LLR(2),LRL(2) 2 ALR(2),LLL(2),LRL(2) ALR LLL 3 4 ARR(1),ALL(2), LLL(2),LRL(2) RRR(0),ALL(2), LLL(2),LRL(2) ARR ALL RRR Tie breaks with the leftmost state in the fringe 40

41 Landmarks! Fact Landmark: a fact that must be true at some state during every plan! Action Landmark: an action that must appear in every valid plan! Disjunctive Action Landmark: a set of actions, one of which must appear in every valid plan! Cost of Disjunctive Action Landmark: the cost of its cheapest action 41

42 Landmark Heuristic! An admissible heuristic! State of the art for optimal planning Basic idea:! Compute a set of disjunctive action landmarks, using the relaxed planing problem! Combine their cost intelligently

43 Relaxed Planning! Relaxed planning: ignore negative effects of an action! Anything achieved cannot be destroyed! Many heuristics operate on the relaxed planning problem:! Original Load(package,vehicle,location)! Preconditions: at(package,location), at(vehicle,location)! Effects: -at(package, location), in(package, vehicle)! Relaxed Load(package,vehicle,location)! Preconditions: at(package,location), at(vehicle,location)! Effects: in(package, vehicle)

44 Landmarks: Example! Remove negative effects (relax )! Propositions: i,a,b,c,g! Initial state {i}. Goal condition {g}! Relaxed Actions:! o 1 [3]: i " a,b! o 2 [4] : i " a,c! o 3 [5] : i " b,c! o 4 [0]: a,b,c " g! An optimal plan:!o 1 o 2 o 4! Cost:!3+4+0 = 7! Landmarks:!W={o 4 } (cost 0)!X={o 1, o 2 } (cost 3)!Y={o 1, o 3 } (cost 3)!Z={o 2, o 3 } (cost 4)!Also: {o 1, o 2, o 3 } (cost 3)

45 Landmarks as Heuristics! Every landmark cost is a lower bound on plan cost (why?)! What if we have a couple of landmarks?! Their max cost is still admissible (= 4)! Sum may not be admissible (=10)!We can do better with hitting sets! W={o 4 } (cost 0) X={o 1, o 2 } (cost 3) Y={o 1, o 3 } (cost 3) Z={o 2, o 3 } (cost 4)

46 Hitting Sets! Hitting set: a set of actions, one of which appears in every landmark! o 1, o 2, o 4 or o 2, o 3, o 4! Hitting set cost: sum of actions costs! Cost (o 1, o 2, o 4 )=7!Cost (o 1, o 2, o 3 )=8! Cost of minimum hitting set is always admissible! (= 7) why?! W={o 4 } (cost 0) X={o 1, o 2 } (cost 3) Y={o 1, o 3 } (cost 3) Z={o 2, o 3 } (cost 4) Cost(o 1 )=3 Cost(o 2 )=4 Cost(o 3 )=5 Cost(o 4 )=0

47 Computing Landmarks 1. Relax the original planning problem! Ignore negative effects 2. Select one precondition for every action 3. Construct the justification graph Nodes = propositions Edges: from chosen precondition to every effect 4. Any cut in this graph is a disjunctive action landmark

48 Pattern database heuristics Example: justification graph Example pcf D: D(o 1 )=D(o 2 )=D(o 3 )=i, D(o 4 )=b o 1 a o 1 [3] : i a, b o 2 [4] : i a, c o 3 [5] : i b, c o 4 [0] : a, b, c g i o 2 o 1 o 3 o 2 b o 4 g o 3 c 48

49 Pattern database heuristics Example: cuts of a justification graph Example Landmark W = {o 4 } (cost 0) o 1 a o 1 [3] : i a, b o 2 [4] : i a, c o 3 [5] : i b, c o 4 [0] : a, b, c g i o 2 o 1 o 3 o 2 b o 4 g o 3 c 49

50 Pattern database heuristics Example: cuts of a justification graph Example Landmark X = {o 1, o 2 } (cost 3) o 1 a o 2 o4 o 1 [3] : i a, b o 2 [4] : i a, c o 3 [5] : i b, c o 4 [0] : a, b, c g i o 1 o 3 o 2 b g o 3 c 50

51 Pattern database heuristics Example: cuts of a justification graph Example Landmark Y = {o 1, o 3 } (cost 3) o 1 a o 1 [3] : i a, b o 2 [4] : i a, c o 3 [5] : i b, c o 4 [0] : a, b, c g i o 2 o 1 o 3 o 2 b o 4 g o 3 c 51

52 Pattern database heuristics Example: cuts of a justification graph Example Landmark Z = {o 2, o 3 } (cost 4) o 1 a o 1 [3] : i a, b o 2 [4] : i a, c o 3 [5] : i b, c o 4 [0] : a, b, c g i o 2 o 1 o 3 o 2 b o 4 g o 3 c 52

53 Beyond Classical Planning! Temporal planning: actions that have duration! Planning under uncertainty:! Unknown initial state, non-deterministic actions, no observations (conformant planning)! Unknown initial state, non-deterministic actions, observations (contingent planning)! Stochastic actions, full observability (Markov Decision Processes! Stochastic actions, partial observability (Partially observable Markov Decision Processes)

54 Summary! Planning is an exciting area of research with many applications! The key to successful planning is the ability to generate a good heuristic function! Heuristics are typically generated by analyzing a simpler problem! Other important techniques are! Domains analysis attempt to detect and exploit properties of a domains, such as symmetry! Pruning techniques methods that allow us to ignore certain branches in the tree, due to symmetry and other reasons

Foundations of Artificial Intelligence

Foundations of Artificial Intelligence Foundations of Artificial Intelligence 37. Automated Planning: Abstraction and Pattern Databases Martin Wehrle Universität Basel May 13, 2016 Planning Heuristics We consider three basic ideas for general

More information

Planning. CPS 570 Ron Parr. Some Actual Planning Applications. Used to fulfill mission objectives in Nasa s Deep Space One (Remote Agent)

Planning. CPS 570 Ron Parr. Some Actual Planning Applications. Used to fulfill mission objectives in Nasa s Deep Space One (Remote Agent) Planning CPS 570 Ron Parr Some Actual Planning Applications Used to fulfill mission objectives in Nasa s Deep Space One (Remote Agent) Particularly important for space operations due to latency Also used

More information

Fast Forward Planning. By J. Hoffmann and B. Nebel

Fast Forward Planning. By J. Hoffmann and B. Nebel Fast Forward Planning By J. Hoffmann and B. Nebel Chronology faster Traditional Optimal Sub-optimal ««1995 Graphplan-based Satplan-based 2000 Heuristic-based Fast Forward (FF) Winner of AIPS2000 Forward-chaining

More information

Foundations of Artificial Intelligence

Foundations of Artificial Intelligence Foundations of Artificial Intelligence 10. Action Planning Solving Logically Specified Problems using a General Problem Solver Joschka Boedecker and Wolfram Burgard and Frank Hutter and Bernhard Nebel

More information

Foundations of Artificial Intelligence

Foundations of Artificial Intelligence Foundations of Artificial Intelligence 39. Automated Planning: Landmark Heuristics Malte Helmert and Gabriele Röger University of Basel May 10, 2017 Automated Planning: Overview Chapter overview: planning

More information

Acknowledgements. Outline

Acknowledgements. Outline Acknowledgements Heuristic Search for Planning Sheila McIlraith University of Toronto Fall 2010 Many of the slides used in today s lecture are modifications of slides developed by Malte Helmert, Bernhard

More information

Heuristic Search for Planning

Heuristic Search for Planning Heuristic Search for Planning Sheila McIlraith University of Toronto Fall 2010 S. McIlraith Heuristic Search for Planning 1 / 50 Acknowledgements Many of the slides used in today s lecture are modifications

More information

Principles of AI Planning. Principles of AI Planning. 7.1 How to obtain a heuristic. 7.2 Relaxed planning tasks. 7.1 How to obtain a heuristic

Principles of AI Planning. Principles of AI Planning. 7.1 How to obtain a heuristic. 7.2 Relaxed planning tasks. 7.1 How to obtain a heuristic Principles of AI Planning June 8th, 2010 7. Planning as search: relaxed planning tasks Principles of AI Planning 7. Planning as search: relaxed planning tasks Malte Helmert and Bernhard Nebel 7.1 How to

More information

Foundations of Artificial Intelligence

Foundations of Artificial Intelligence Foundations of Artificial Intelligence 11. Action Planning Solving Logically Specified Problems using a General Problem Solver Joschka Boedecker and Wolfram Burgard and Bernhard Nebel Albert-Ludwigs-Universität

More information

Searching with Partial Information

Searching with Partial Information Searching with Partial Information Above we (unrealistically) assumed that the environment is fully observable and deterministic Moreover, we assumed that the agent knows what the effects of each action

More information

Set 9: Planning Classical Planning Systems. ICS 271 Fall 2014

Set 9: Planning Classical Planning Systems. ICS 271 Fall 2014 Set 9: Planning Classical Planning Systems ICS 271 Fall 2014 Planning environments Classical Planning: Outline: Planning Situation calculus PDDL: Planning domain definition language STRIPS Planning Planning

More information

Implicit Abstraction Heuristics for Cost-Optimal Planning

Implicit Abstraction Heuristics for Cost-Optimal Planning Abstraction for Cost-Optimal Planning Michael Katz Advisor: Carmel Domshlak Technion - Israel Institute of Technology Planning Language Classical Planning in SAS + Planning task is 5-tuple V, A, C, s 0,

More information

Set 9: Planning Classical Planning Systems. ICS 271 Fall 2013

Set 9: Planning Classical Planning Systems. ICS 271 Fall 2013 Set 9: Planning Classical Planning Systems ICS 271 Fall 2013 Outline: Planning Classical Planning: Situation calculus PDDL: Planning domain definition language STRIPS Planning Planning graphs Readings:

More information

Creating Admissible Heuristic Functions: The General Relaxation Principle and Delete Relaxation

Creating Admissible Heuristic Functions: The General Relaxation Principle and Delete Relaxation Creating Admissible Heuristic Functions: The General Relaxation Principle and Delete Relaxation Relaxed Planning Problems 74 A classical planning problem P has a set of solutions Solutions(P) = { π : π

More information

Planning and Optimization

Planning and Optimization Planning and Optimization F3. Post-hoc Optimization & Operator Counting Malte Helmert and Gabriele Röger Universität Basel December 6, 2017 Post-hoc Optimization Heuristic Content of this Course: Heuristic

More information

Intelligent Agents. State-Space Planning. Ute Schmid. Cognitive Systems, Applied Computer Science, Bamberg University. last change: 14.

Intelligent Agents. State-Space Planning. Ute Schmid. Cognitive Systems, Applied Computer Science, Bamberg University. last change: 14. Intelligent Agents State-Space Planning Ute Schmid Cognitive Systems, Applied Computer Science, Bamberg University last change: 14. April 2016 U. Schmid (CogSys) Intelligent Agents last change: 14. April

More information

Searching. Assume goal- or utilitybased. Next task to achieve is to determine the best path to the goal

Searching. Assume goal- or utilitybased. Next task to achieve is to determine the best path to the goal Searching Assume goal- or utilitybased agents: state information ability to perform actions goals to achieve Next task to achieve is to determine the best path to the goal CSC384 Lecture Slides Steve Engels,

More information

36.1 Relaxed Planning Graphs

36.1 Relaxed Planning Graphs Foundations of Artificial Intelligence ay 9, 26 6. Automated Planning: Delete Relaxation Heuristics Foundations of Artificial Intelligence 6. Automated Planning: Delete Relaxation Heuristics artin Wehrle

More information

INF Kunstig intelligens. Agents That Plan. Roar Fjellheim. INF5390-AI-06 Agents That Plan 1

INF Kunstig intelligens. Agents That Plan. Roar Fjellheim. INF5390-AI-06 Agents That Plan 1 INF5390 - Kunstig intelligens Agents That Plan Roar Fjellheim INF5390-AI-06 Agents That Plan 1 Outline Planning agents Plan representation State-space search Planning graphs GRAPHPLAN algorithm Partial-order

More information

A deterministic action is a partial function from states to states. It is partial because not every action can be carried out in every state

A deterministic action is a partial function from states to states. It is partial because not every action can be carried out in every state CmSc310 Artificial Intelligence Classical Planning 1. Introduction Planning is about how an agent achieves its goals. To achieve anything but the simplest goals, an agent must reason about its future.

More information

Probabilistic Belief. Adversarial Search. Heuristic Search. Planning. Probabilistic Reasoning. CSPs. Learning CS121

Probabilistic Belief. Adversarial Search. Heuristic Search. Planning. Probabilistic Reasoning. CSPs. Learning CS121 CS121 Heuristic Search Planning CSPs Adversarial Search Probabilistic Reasoning Probabilistic Belief Learning Heuristic Search First, you need to formulate your situation as a Search Problem What is a

More information

Domain-Dependent Heuristics and Tie-Breakers: Topics in Automated Planning

Domain-Dependent Heuristics and Tie-Breakers: Topics in Automated Planning Domain-Dependent Heuristics and Tie-Breakers: Topics in Automated Planning Augusto B. Corrêa, André G. Pereira, Marcus Ritt 1 Instituto de Informática Universidade Federal do Rio Grande do Sul (UFRGS)

More information

Preferences, Planning, and Control

Preferences, Planning, and Control Department of Computer Science Ben-Gurion University September 18, 2008 Outline 1 Why preferences? 2 A graphical preference model 3 Planning and Preferences 1 Planning with Preferences 2 Graphical Models

More information

A lookahead strategy for heuristic search planning

A lookahead strategy for heuristic search planning A lookahead strategy for heuristic search planning Vincent Vidal IRIT Université Paul Sabatier 118 route de Narbonne 31062 Toulouse Cedex 04, France email: vvidal@irit.fr Technical report IRIT/2002-35-R

More information

Conformant Planning via Heuristic Forward Search: A New Approach

Conformant Planning via Heuristic Forward Search: A New Approach Conformant Planning via Heuristic Forward Search: A New Approach Ronen I. Brafman Dept. of Computer Science Ben-Gurion University Beer-Sheva 84105, Israel Jörg Hoffmann Dept. of Computer Science Albert-Ludwigs

More information

Extending PDDL for Hierarchical Planning and Topological Abstraction

Extending PDDL for Hierarchical Planning and Topological Abstraction Extending PDDL for Hierarchical Planning and Topological Abstraction Adi Botea and Martin Müller and Jonathan Schaeffer Department of Computing Science, University of Alberta Edmonton, Alberta, Canada

More information

Planning: Wrap up CSP Planning Logic: Intro

Planning: Wrap up CSP Planning Logic: Intro Planning: Wrap up CSP Planning Logic: Intro Alan Mackworth UBC CS 322 Planning 3 ebruary 25, 2013 Textbook 8.4, 5.1 GAC vs. AC GAC = Generalized Arc Consistency (for k-ary constraints) AC = Arc Consistency

More information

CMU-Q Lecture 6: Planning Graph GRAPHPLAN. Teacher: Gianni A. Di Caro

CMU-Q Lecture 6: Planning Graph GRAPHPLAN. Teacher: Gianni A. Di Caro CMU-Q 15-381 Lecture 6: Planning Graph GRAPHPLAN Teacher: Gianni A. Di Caro PLANNING GRAPHS Graph-based data structure representing a polynomial-size/time approximation of the exponential search tree Can

More information

CIS 192: Artificial Intelligence. Search and Constraint Satisfaction Alex Frias Nov. 30 th

CIS 192: Artificial Intelligence. Search and Constraint Satisfaction Alex Frias Nov. 30 th CIS 192: Artificial Intelligence Search and Constraint Satisfaction Alex Frias Nov. 30 th What is AI? Designing computer programs to complete tasks that are thought to require intelligence 4 categories

More information

Validating Plans with Durative Actions via Integrating Boolean and Numerical Constraints

Validating Plans with Durative Actions via Integrating Boolean and Numerical Constraints Validating Plans with Durative Actions via Integrating Boolean and Numerical Constraints Roman Barták Charles University in Prague, Faculty of Mathematics and Physics Institute for Theoretical Computer

More information

Monte Carlo Tree Search PAH 2015

Monte Carlo Tree Search PAH 2015 Monte Carlo Tree Search PAH 2015 MCTS animation and RAVE slides by Michèle Sebag and Romaric Gaudel Markov Decision Processes (MDPs) main formal model Π = S, A, D, T, R states finite set of states of the

More information

Planning & Decision-making in Robotics Case Study: Planning for Autonomous Driving

Planning & Decision-making in Robotics Case Study: Planning for Autonomous Driving 16-782 Planning & Decision-making in Robotics Case Study: Planning for Autonomous Driving Maxim Likhachev Robotics Institute Carnegie Mellon University Typical Planning Architecture for Autonomous Vehicle

More information

Metis: Arming Fast Downward with Pruning and Incremental Computation

Metis: Arming Fast Downward with Pruning and Incremental Computation Metis: Arming Fast Downward with Pruning and Incremental Computation Yusra Alkhazraji University of Freiburg, Germany alkhazry@informatik.uni-freiburg.de Florian Pommerening University of Basel, Switzerland

More information

3 SOLVING PROBLEMS BY SEARCHING

3 SOLVING PROBLEMS BY SEARCHING 48 3 SOLVING PROBLEMS BY SEARCHING A goal-based agent aims at solving problems by performing actions that lead to desirable states Let us first consider the uninformed situation in which the agent is not

More information

AltAlt p : Online Parallelization of Plans with Heuristic State Search

AltAlt p : Online Parallelization of Plans with Heuristic State Search ONLINE PARALLELIZATION OF PLANS WITH HEURISTIC STATE SEARCH AltAlt p : Online Parallelization of Plans with Heuristic State Search Romeo Sanchez Nigenda Subbarao Kambhampati DEPARTMENT OF COMPUTER SCIENCE

More information

INTRODUCTION TO HEURISTIC SEARCH

INTRODUCTION TO HEURISTIC SEARCH INTRODUCTION TO HEURISTIC SEARCH What is heuristic search? Given a problem in which we must make a series of decisions, determine the sequence of decisions which provably optimizes some criterion. What

More information

15-780: Graduate AI Homework Assignment #2 Solutions

15-780: Graduate AI Homework Assignment #2 Solutions 15-780: Graduate AI Homework Assignment #2 Solutions Out: February 12, 2015 Due: February 25, 2015 Collaboration Policy: You may discuss the problems with others, but you must write all code and your writeup

More information

BAYESIAN NETWORKS STRUCTURE LEARNING

BAYESIAN NETWORKS STRUCTURE LEARNING BAYESIAN NETWORKS STRUCTURE LEARNING Xiannian Fan Uncertainty Reasoning Lab (URL) Department of Computer Science Queens College/City University of New York http://url.cs.qc.cuny.edu 1/52 Overview : Bayesian

More information

The Fast Downward Planning System

The Fast Downward Planning System Journal of Artificial Intelligence Research 26 (2006) 191 246 Submitted 01/05; published 07/06 The Fast Downward Planning System Malte Helmert Institut für Informatik Albert-Ludwigs-Universität Freiburg

More information

4 INFORMED SEARCH AND EXPLORATION. 4.1 Heuristic Search Strategies

4 INFORMED SEARCH AND EXPLORATION. 4.1 Heuristic Search Strategies 55 4 INFORMED SEARCH AND EXPLORATION We now consider informed search that uses problem-specific knowledge beyond the definition of the problem itself This information helps to find solutions more efficiently

More information

Today s s lecture. Lecture 3: Search - 2. Problem Solving by Search. Agent vs. Conventional AI View. Victor R. Lesser. CMPSCI 683 Fall 2004

Today s s lecture. Lecture 3: Search - 2. Problem Solving by Search. Agent vs. Conventional AI View. Victor R. Lesser. CMPSCI 683 Fall 2004 Today s s lecture Search and Agents Material at the end of last lecture Lecture 3: Search - 2 Victor R. Lesser CMPSCI 683 Fall 2004 Continuation of Simple Search The use of background knowledge to accelerate

More information

A Case-Based Approach to Heuristic Planning

A Case-Based Approach to Heuristic Planning Noname manuscript No. (will be inserted by the editor) A Case-Based Approach to Heuristic Planning Tomás de la Rosa Angel García-Olaya Daniel Borrajo the date of receipt and acceptance should be inserted

More information

Benchmarking of Automated Planners: which Planner suits you best?

Benchmarking of Automated Planners: which Planner suits you best? Benchmarking of Automated Planners: which Planner suits you best? Wim Florijn University of Twente P.O. Box 217, 7500AE Enschede The Netherlands w.j.florijn@student.utwente.nl ABSTRACT Benchmarking provides

More information

Classical Planning Problems: Representation Languages

Classical Planning Problems: Representation Languages jonas.kvarnstrom@liu.se 2017 Classical Planning Problems: Representation Languages History: 1959 3 The language of Artificial Intelligence was/is logic First-order, second-order, modal, 1959: General

More information

Planning Graph Heuristics for Belief Space Search

Planning Graph Heuristics for Belief Space Search Journal of Artificial Intelligence Research () Submitted 06/?/05; published Planning Graph Heuristics for Belief Space Search Daniel Bryce Subbarao Kambhampati Department of Computer Science and Engineering

More information

Heuristic (Informed) Search

Heuristic (Informed) Search Heuristic (Informed) Search (Where we try to choose smartly) R&N: Chap. 4, Sect. 4.1 3 1 Recall that the ordering of FRINGE defines the search strategy Search Algorithm #2 SEARCH#2 1. INSERT(initial-node,FRINGE)

More information

Automated Service Composition using Heuristic Search

Automated Service Composition using Heuristic Search Automated Service Composition using Heuristic Search Harald Meyer and Mathias Weske Hasso-Plattner-Institute for IT-Systems-Engineering at the University of Potsdam Prof.-Dr.-Helmert-Strasse 2-3, 14482

More information

Informed (Heuristic) Search. Idea: be smart about what paths to try.

Informed (Heuristic) Search. Idea: be smart about what paths to try. Informed (Heuristic) Search Idea: be smart about what paths to try. 1 Blind Search vs. Informed Search What s the difference? How do we formally specify this? A node is selected for expansion based on

More information

Joint Entity Resolution

Joint Entity Resolution Joint Entity Resolution Steven Euijong Whang, Hector Garcia-Molina Computer Science Department, Stanford University 353 Serra Mall, Stanford, CA 94305, USA {swhang, hector}@cs.stanford.edu No Institute

More information

This article was published in an Elsevier journal. The attached copy is furnished to the author for non-commercial research and education use, including for instruction at the author s institution, sharing

More information

Search. (Textbook Chpt ) Computer Science cpsc322, Lecture 2. May, 10, CPSC 322, Lecture 2 Slide 1

Search. (Textbook Chpt ) Computer Science cpsc322, Lecture 2. May, 10, CPSC 322, Lecture 2 Slide 1 Search Computer Science cpsc322, Lecture 2 (Textbook Chpt 3.0-3.4) May, 10, 2012 CPSC 322, Lecture 2 Slide 1 Colored Cards You need to have 4 colored index cards Come and get them from me if you still

More information

Temporal Planning with a Non-Temporal Planner

Temporal Planning with a Non-Temporal Planner Temporal Planning with a Non-Temporal Planner Keith Halsey Department of Computer Science, University of Durham, UK keith.halsey@durham.ac.uk Introduction This extended abstract looks at the work that

More information

Planning. Introduction

Planning. Introduction Introduction vs. Problem-Solving Representation in Systems Situation Calculus The Frame Problem STRIPS representation language Blocks World with State-Space Search Progression Algorithms Regression Algorithms

More information

Artificial Intelligence

Artificial Intelligence Artificial Intelligence Other models of interactive domains Marc Toussaint University of Stuttgart Winter 2018/19 Basic Taxonomy of domain models Other models of interactive domains Basic Taxonomy of domain

More information

Search Algorithms for Planning

Search Algorithms for Planning Search Algorithms for Planning Sheila McIlraith University of Toronto Fall 2010 S. McIlraith Search Algorithms 1 / 50 Acknowledgements Many of the slides used in today s lecture are modifications of slides

More information

Feature Subset Selection using Clusters & Informed Search. Team 3

Feature Subset Selection using Clusters & Informed Search. Team 3 Feature Subset Selection using Clusters & Informed Search Team 3 THE PROBLEM [This text box to be deleted before presentation Here I will be discussing exactly what the prob Is (classification based on

More information

Artificial Intelligence

Artificial Intelligence Artificial Intelligence Search Marc Toussaint University of Stuttgart Winter 2015/16 (slides based on Stuart Russell s AI course) Outline Problem formulation & examples Basic search algorithms 2/100 Example:

More information

Classical Planning. CS 486/686: Introduction to Artificial Intelligence Winter 2016

Classical Planning. CS 486/686: Introduction to Artificial Intelligence Winter 2016 Classical Planning CS 486/686: Introduction to Artificial Intelligence Winter 2016 1 Classical Planning A plan is a collection of actions for performing some task (reaching some goal) If we have a robot

More information

Researching Heuristic Functions to Detect and Avoid Dead Ends in Action Planning

Researching Heuristic Functions to Detect and Avoid Dead Ends in Action Planning Otto-Friedrich-University Bamberg Master Thesis Researching Heuristic Functions to Detect and Avoid Dead Ends in Action Planning Author: Christian Reißner Supervisor: Prof. Dr. Ute Schmid with Michael

More information

Navigation and Metric Path Planning

Navigation and Metric Path Planning Navigation and Metric Path Planning October 4, 2011 Minerva tour guide robot (CMU): Gave tours in Smithsonian s National Museum of History Example of Minerva s occupancy map used for navigation Objectives

More information

Scalable Web Service Composition with Partial Matches

Scalable Web Service Composition with Partial Matches Scalable Web Service Composition with Partial Matches Adina Sirbu and Jörg Hoffmann Digital Enterprise Research Institute (DERI) University of Innsbruck, Austria firstname.lastname@deri.org Abstract. We

More information

Simultaneous Heuristic Search for Conjunctive Subgoals

Simultaneous Heuristic Search for Conjunctive Subgoals Simultaneous Heuristic Search for Conjunctive Subgoals Lin Zhu and Robert Givan Electrical and Computer Engineering, Purdue University, West Lafayette IN 4797 USA {lzhu, givan}@purdue.edu Abstract We study

More information

Principles of Autonomy and Decision Making

Principles of Autonomy and Decision Making Massachusetts Institute of Technology 16.410-13 Principles of Autonomy and Decision Making Problem Set #6 (distributed 10/22/04). Paper solutions are due no later than 5pm on Friday, 10/29/04. Please give

More information

Abstraction. Today s Lecture. Problem Formalization Issues. Lecture 3: Search 2

Abstraction. Today s Lecture. Problem Formalization Issues. Lecture 3: Search 2 Lecture 3: Search 2 Victor R. Lesser CMPSCI 683 Fall 2010 Today s Lecture Finish off Introductory Material on Search Brief Review of Blind Search How to use heuristics (domain knowledge) in order to accelerate

More information

Flexible Execution of Plans with Choice and Uncertainty Patrick R Conrad and Brian C Williams

Flexible Execution of Plans with Choice and Uncertainty Patrick R Conrad and Brian C Williams Computer Science and Artificial Intelligence Laboratory Technical Report MIT-CSAIL-TR-2011-002 January 15, 2011 Flexible Execution of Plans with Choice and Uncertainty Patrick R Conrad and Brian C Williams

More information

DFS. Depth-limited Search

DFS. Depth-limited Search DFS Completeness? No, fails in infinite depth spaces or spaces with loops Yes, assuming state space finite. Time complexity? O(b m ), terrible if m is much bigger than d. can do well if lots of goals Space

More information

Chapter 3 Complexity of Classical Planning

Chapter 3 Complexity of Classical Planning Lecture slides for Automated Planning: Theory and Practice Chapter 3 Complexity of Classical Planning Dana S. Nau CMSC 722, AI Planning University of Maryland, Spring 2008 Licensed under the Creative Commons

More information

Exam Topics. Search in Discrete State Spaces. What is intelligence? Adversarial Search. Which Algorithm? 6/1/2012

Exam Topics. Search in Discrete State Spaces. What is intelligence? Adversarial Search. Which Algorithm? 6/1/2012 Exam Topics Artificial Intelligence Recap & Expectation Maximization CSE 473 Dan Weld BFS, DFS, UCS, A* (tree and graph) Completeness and Optimality Heuristics: admissibility and consistency CSPs Constraint

More information

Planning and Acting. CITS3001 Algorithms, Agents and Artificial Intelligence. 2018, Semester 2

Planning and Acting. CITS3001 Algorithms, Agents and Artificial Intelligence. 2018, Semester 2 Planning and Acting CITS3001 Algorithms, Agents and Artificial Intelligence Tim French School of Computer Science and Software Engineering The University of Western Australia 2018, Semester 2 Summary We

More information

The exam is closed book, closed calculator, and closed notes except your one-page crib sheet.

The exam is closed book, closed calculator, and closed notes except your one-page crib sheet. CS Summer Introduction to Artificial Intelligence Midterm You have approximately minutes. The exam is closed book, closed calculator, and closed notes except your one-page crib sheet. Mark your answers

More information

Symbolic LAO* Search for Factored Markov Decision Processes

Symbolic LAO* Search for Factored Markov Decision Processes Symbolic LAO* Search for Factored Markov Decision Processes Zhengzhu Feng Computer Science Department University of Massachusetts Amherst MA 01003 Eric A. Hansen Computer Science Department Mississippi

More information

Accelerated Raytracing

Accelerated Raytracing Accelerated Raytracing Why is Acceleration Important? Vanilla ray tracing is really slow! mxm pixels, kxk supersampling, n primitives, average ray path length of d, l lights, 2 recursive ray casts per

More information

Computing Applicability Conditions for Plans with Loops

Computing Applicability Conditions for Plans with Loops Abacus Programs Translation Examples Computing Applicability Conditions for Plans with Loops Siddharth Srivastava Neil Immerman Shlomo Zilberstein Department of Computer Science, University of Massachusetts

More information

Markov Decision Processes. (Slides from Mausam)

Markov Decision Processes. (Slides from Mausam) Markov Decision Processes (Slides from Mausam) Machine Learning Operations Research Graph Theory Control Theory Markov Decision Process Economics Robotics Artificial Intelligence Neuroscience /Psychology

More information

Intelligent Agents. Planning Graphs - The Graph Plan Algorithm. Ute Schmid. Cognitive Systems, Applied Computer Science, Bamberg University

Intelligent Agents. Planning Graphs - The Graph Plan Algorithm. Ute Schmid. Cognitive Systems, Applied Computer Science, Bamberg University Intelligent Agents Planning Graphs - The Graph Plan Algorithm Ute Schmid Cognitive Systems, Applied Computer Science, Bamberg University last change: July 9, 2015 U. Schmid (CogSys) Intelligent Agents

More information

Heuristic Search. Heuristic Search. Heuristic Search. CSE 3401: Intro to AI & LP Informed Search

Heuristic Search. Heuristic Search. Heuristic Search. CSE 3401: Intro to AI & LP Informed Search CSE 3401: Intro to AI & LP Informed Search Heuristic Search. Required Readings: Chapter 3, Sections 5 and 6, and Chapter 4, Section 1. In uninformed search, we don t try to evaluate which of the nodes

More information

Slides credited from Dr. David Silver & Hung-Yi Lee

Slides credited from Dr. David Silver & Hung-Yi Lee Slides credited from Dr. David Silver & Hung-Yi Lee Review Reinforcement Learning 2 Reinforcement Learning RL is a general purpose framework for decision making RL is for an agent with the capacity to

More information

Informed Search Methods

Informed Search Methods Informed Search Methods How can we improve searching strategy by using intelligence? Map example: Heuristic: Expand those nodes closest in as the crow flies distance to goal 8-puzzle: Heuristic: Expand

More information

Artificial Intelligence. Planning

Artificial Intelligence. Planning Artificial Intelligence Planning Planning Planning agent Similar to previous problem solving agents Constructs plans that achieve its goals, then executes them Differs in way it represents and searches

More information

Introduction to System Design

Introduction to System Design Introduction to System Design Software Requirements and Design CITS 4401 Lecture 8 System Design is a creative process no cook book solutions goal driven we create a design for solving some problem constraint

More information

Heuristic (Informed) Search

Heuristic (Informed) Search Heuristic (Informed) Search (Where we try to choose smartly) R&N: Chap., Sect..1 3 1 Search Algorithm #2 SEARCH#2 1. INSERT(initial-node,Open-List) 2. Repeat: a. If empty(open-list) then return failure

More information

Problem Solving & Heuristic Search

Problem Solving & Heuristic Search 190.08 Artificial 2016-Spring Problem Solving & Heuristic Search Byoung-Tak Zhang School of Computer Science and Engineering Seoul National University 190.08 Artificial (2016-Spring) http://www.cs.duke.edu/courses/fall08/cps270/

More information

Discussion Section. Matthew Tong

Discussion Section. Matthew Tong Discussion Section Matthew Tong HW #1 1) Define Intelligence AI Agent 2) Intractable/unsolvable => no AI Big difference between the two Intractable leads open possibility of better machines or algorithms

More information

Fast and Informed Action Selection for Planning with Sensing

Fast and Informed Action Selection for Planning with Sensing Fast and Informed Action Selection for Planning with Sensing Alexandre Albore 1, Héctor Palacios 1, and Hector Geffner 2 1 Universitat Pompeu Fabra Passeig de Circumvalació 8 08003 Barcelona Spain 2 ICREA

More information

Structure and Complexity in Planning with Unary Operators

Structure and Complexity in Planning with Unary Operators Structure and Complexity in Planning with Unary Operators Carmel Domshlak and Ronen I Brafman ½ Abstract In this paper we study the complexity of STRIPS planning when operators have a single effect In

More information

Stuck in Traffic (SiT) Attacks

Stuck in Traffic (SiT) Attacks Stuck in Traffic (SiT) Attacks Mina Guirguis Texas State University Joint work with George Atia Traffic 2 Intelligent Transportation Systems V2X communication enable drivers to make better decisions: Avoiding

More information

Chapter 1 Introduc0on

Chapter 1 Introduc0on Lecture slides for Automated Planning: Theory and Practice Chapter 1 Introduc0on Dana S. Nau University of Maryland Updated 3:03 PM December 14, 2013 1 Some Dic0onary Defini0ons of Plan plan n. 1. A scheme,

More information

State Agnostic Planning Graphs: Deterministic, Non-Deterministic, and Probabilistic Planning

State Agnostic Planning Graphs: Deterministic, Non-Deterministic, and Probabilistic Planning State Agnostic Planning Graphs: Deterministic, Non-Deterministic, and Probabilistic Planning Daniel Bryce, a William Cushing, b and Subbarao Kambhampati b a Utah State University, Department of Computer

More information

Artificial Intelligence Informed search. Peter Antal

Artificial Intelligence Informed search. Peter Antal Artificial Intelligence Informed search Peter Antal antal@mit.bme.hu 1 Informed = use problem-specific knowledge Which search strategies? Best-first search and its variants Heuristic functions? How to

More information

A weighted CSP approach to cost-optimal planning

A weighted CSP approach to cost-optimal planning 1 A weighted CSP approach to cost-optimal planning Martin C. Cooper, Marie de Roquemaurel and Pierre Régnier IRIT, University of Toulouse III, 118 route de Narbonne, 31062 Toulouse, France E-mail: {cooper,deroquemaurel,regnier}@irit.fr

More information

ˆ The exam is closed book, closed calculator, and closed notes except your one-page crib sheet.

ˆ The exam is closed book, closed calculator, and closed notes except your one-page crib sheet. CS Summer Introduction to Artificial Intelligence Midterm ˆ You have approximately minutes. ˆ The exam is closed book, closed calculator, and closed notes except your one-page crib sheet. ˆ Mark your answers

More information

Planning with Goal Preferences and Constraints

Planning with Goal Preferences and Constraints Planning with Goal Preferences and Constraints Ronen I. Brafman NASA Ames Research Center and Computer Science Dept. Stanford University brafman@email.arc.nasa.gov Yuri Chernyavsky Computer Science Department

More information

ScienceDirect. Plan Restructuring in Multi Agent Planning

ScienceDirect. Plan Restructuring in Multi Agent Planning Available online at www.sciencedirect.com ScienceDirect Procedia Computer Science 46 (2015 ) 396 401 International Conference on Information and Communication Technologies (ICICT 2014) Plan Restructuring

More information

Planning for Markov Decision Processes with Sparse Stochasticity

Planning for Markov Decision Processes with Sparse Stochasticity Planning for Markov Decision Processes with Sparse Stochasticity Maxim Likhachev Geoff Gordon Sebastian Thrun School of Computer Science School of Computer Science Dept. of Computer Science Carnegie Mellon

More information

Lecture 4: Search 3. Victor R. Lesser. CMPSCI 683 Fall 2010

Lecture 4: Search 3. Victor R. Lesser. CMPSCI 683 Fall 2010 Lecture 4: Search 3 Victor R. Lesser CMPSCI 683 Fall 2010 First Homework 1 st Programming Assignment 2 separate parts (homeworks) First part due on (9/27) at 5pm Second part due on 10/13 at 5pm Send homework

More information

Advanced Operations Research Prof. G. Srinivasan Department of Management Studies Indian Institute of Technology, Madras

Advanced Operations Research Prof. G. Srinivasan Department of Management Studies Indian Institute of Technology, Madras Advanced Operations Research Prof. G. Srinivasan Department of Management Studies Indian Institute of Technology, Madras Lecture - 35 Quadratic Programming In this lecture, we continue our discussion on

More information

Chapter 2 Classical algorithms in Search and Relaxation

Chapter 2 Classical algorithms in Search and Relaxation Chapter 2 Classical algorithms in Search and Relaxation Chapter 2 overviews topics on the typical problems, data structures, and algorithms for inference in hierarchical and flat representations. Part

More information

Strong Controllability of Disjunctive Temporal Problems with Uncertainty

Strong Controllability of Disjunctive Temporal Problems with Uncertainty Strong Controllability of Disjunctive Temporal Problems with Uncertainty Bart Peintner 1, Kristen Brent Venable 2, and Neil Yorke-Smith 1 1 Artificial Intelligence Center, SRI International. {peintner,nysmith}@ai.sri.com

More information

Machine Learning for Software Engineering

Machine Learning for Software Engineering Machine Learning for Software Engineering Introduction and Motivation Prof. Dr.-Ing. Norbert Siegmund Intelligent Software Systems 1 2 Organizational Stuff Lectures: Tuesday 11:00 12:30 in room SR015 Cover

More information

Task-Oriented Planning for Manipulating Articulated Mechanisms under Model Uncertainty. Venkatraman Narayanan and Maxim Likhachev

Task-Oriented Planning for Manipulating Articulated Mechanisms under Model Uncertainty. Venkatraman Narayanan and Maxim Likhachev Task-Oriented Planning for Manipulating Articulated Mechanisms under Model Uncertainty Venkatraman Narayanan and Maxim Likhachev Motivation Motivation A lot of household objects are ARTICULATED A lot of

More information