Planning in Artificial Intelligence: Sequence of Actions to Achieve a Goal, Lecture notes of Artificial Intelligence

Download Planning in Artificial Intelligence: Sequence of Actions to Achieve a Goal and more Lecture notes Artificial Intelligence in PDF only on Docsity!

Planning

AI Planning

• The task of coming up with a sequence of actions that will achieve a goal is called planning.
• Until recently, AI planning was essentially a theoretical endeavor: It‘s now becoming useful in industrial applications
• Example application areas
  • design & manufacturing
  • military operations & logistics
  • games
  • space exploration -Proof Planning in Mathematics
  • Speech and Dialog Planning

Notions

• Plan sequence of (actions) transforming the initial state into a final state
• Operators representation of actions Room 2
• Planner algorithm that generates a plan from a (partial) description of initial and final state and from a specification of operators

Go to the basket

Classical Planning Environments

• Fully observable
• Deterministic outcome
• Finite steps
• Static (change happens only due to action)
• Discrete in time, action, objects and effects

The classical Planning Framework

Example: Travel Scenario

Two phases: Plan Generation and Plan Execution

Given: Initial state, Goal State and a Set of Operators

Task: Find planning operators which transform initial state into goal state

Describe states in a formal language like Predicate Logic

• A Plan is a sequence of operators that transforms the initial state into the goal state

• Initial State- At Home
• Goal State- Get milk, banana and a cordless drill.
• Depending on the set of actions… branching factor could be huge.
• Heuristic functions can guide the search of states, but cannot eliminate states.

• Problems with a search agent

Too many actions and too many states to

consider

 Heuristic functions can only choose among

states and can’t eliminate actions from

consideration; so which action should be

taken?

 Agent is forced to consider actions starting

from the initial state.

From problem solving to planning

• They are different in representing goals, states, and actions, and in ways of constructing action sequences.
• For a problem-solving agent,

 Actions - generate successor state descriptions

 State representations for successor generation, heuristic eval function, and goal testing

 Goals used in the goal test and the heuristic function

Classical planning environments

• Fully observable
• Deterministic outcome
• Finite steps
• Static (change happens only due to action)
• Discrete in time, action, objects, and effects

STRIPS

• Before starting STRIPS….How is this different from problem-solving via search?
• states, goals, and actions use representations in some formal language we can look inside actions: - with search, actions defined only by state transitions - with planning, describe actions by:  preconditions: what must be true before an action can be performed  effects (postconditions): what must be true (what has changed) after action is performed - this allows planning systems to reason about interaction between different actions
• we can consider actions in any order - with search, only consider continuous sequences of actions from initial state to goal - with planning: we can add steps incrementally, regardless of their final position in the sequence

STRIPS

• STRIPS: Stanford Research

Institute Planning System (1970)

• a restricted language

(essentially propositional logic) used in most classical planners

• First major planning system -

planner for the Shakey robot

Basic Representation

• STRIPS language for efficient planning

 States are conjunctions of function-free ground literals At(P1,JFK)^At(P2,SFO)^Plane(P1)^Plane(P2)^..

 Goals are conjunctions of literals, can contain variables At(C1,JFK)^At(C2,SFO)

 Implicit representations of states in planning - only changes are tracked.

 Actions in STRIPS have three components: action, precondition, and effect

STRIPS (2)

• An action is applicable in any state that satisfies the precondition; else the action has no effect
• An example about applicable application At(P1,JFK)^At(P2,SFO)^Plane(P1)^Plane(P2)^Airport(JFK) ^Airport(SFO), Is the action Fly applicable? After flying, we have At(P1,SFO)^At(P2,SFO)^Plane(P1)^Plane(P2)^Airport(JFK) ^Airport(SFO), Is Fly applicable?
• STRIPS assumption: Every literal not mentioned in the effect remains unchanged The Frame problem is avoided
• The solution for a planning problem is a sequence of actions, starting in the initial state, results in a goal state

Another Example: Representing

States of the World

• State: a consistent assignment of TRUE or

FALSE to every literal in the universe

• State description:- a set of ground literals that

are all taken to be TRUE

 The negation of these literals are taken to be false  Truth values of other ground literals are unknown

• Note: in standard STRIPS, a state is restricted

to contain only positive literals