1. www.sti-innsbruck.at © Copyright 2008 STI INNSBRUCK www.sti-innsbruck.at Intelligent Systems Intelligent Agents – Lecture 9 Prof. Dieter Fensel (& Francois Scharffe)

2. www.sti-innsbruck.at Agenda Motivation – Agents and environments – Rationality Technical solution – PEAS (Performance measure, Environment, Actuators, Sensors)‏ – Environment types – Agent types Illustration Extensions Conclusion 2

3. www.sti-innsbruck.at Agents An agent is anything that can be viewed as perceiving its environment through sensors and acting upon that environment through actuators Human agent: eyes, ears, and other organs for sensors; hands, legs, mouth, and other body parts for actuators Robotic agent: cameras and infrared range finders for sensors; various motors for actuators 3

4. www.sti-innsbruck.at Agents and environments The agent function maps from percept histories to actions: [f: P*  A ] The agent program runs on the physical architecture to produce f agent = architecture + program 4

5. www.sti-innsbruck.at Vacuum-cleaner world Percepts: location and contents, e.g., [A,Dirty] Actions: Left, Right, Suck, NoOp http://www.dylangreene.com/posts/442 (Dyson DC 06 Robotic Vacuum Cleaner) http://www.dylangreene.com/posts/442 5

6. www.sti-innsbruck.at A vacuum-cleaner agent function Reflex-Vacuum-Agent( [location,status]) returns an action if status = Dirty then return Clean else if location = A then return Right else if location = B then return Left 6 Clean[A;Clean], [A;Dirty] Right[A;Clean], [A;Clean] Clean[B;Dirty] Left[B;Clean] Clean[A;Dirty] Right[A;Clean] ActionPercept sequence

7. www.sti-innsbruck.at Rational agents An agent should strive to "do the right thing", based on what it can perceive and the actions it can perform. The right action is the one that will cause the agent to be most successful Performance measure: An objective criterion for success of an agent's behavior E.g., performance measure of a vacuum-cleaner agent could be amount of dirt cleaned up, amount of time taken, amount of electricity consumed, amount of noise generated, etc. 7

8. www.sti-innsbruck.at Rational agents Rational Agent: For each possible percept sequence, a rational agent should select an action that is expected to – maximize its performance measure, – given the evidence provided by the percept sequence, and – whatever built-in knowledge the agent has. 8

9. www.sti-innsbruck.at Rational agents Rationality is distinct from omniscience (all-knowing with infinite knowledge)‏ Agents can perform actions in order to modify future percepts so as to obtain useful information (information gathering, exploration)‏ An agent is autonomous if its behavior is determined by its own experience (with ability to learn and adapt)‏ 9

10. www.sti-innsbruck.at PEAS PEAS: Performance measure, Environment, Actuators, Sensors One must first specify the setting for intelligent agent design Consider, e.g., the task of designing an automated taxi driver: – Performance measure – Environment – Actuators – Sensors 10

11. www.sti-innsbruck.at PEAS – Example 1 Agent: Automated taxi driver: – Performance measure: Safe, fast, legal, comfortable trip, maximize profits – Environment: Roads, other traffic, pedestrians, customers – Actuators: Steering wheel, accelerator, brake, signal, horn – Sensors: Cameras, sonar, speedometer, GPS, odometer, taximeter engine sensors, keyboard 11

12. www.sti-innsbruck.at PEAS – Example 2 Agent: Medical diagnosis system: – Performance measure: Healthy patient, minimize costs, lawsuits – Environment: Patient, hospital, staff – Actuators: Screen display (questions, tests, diagnoses, treatments, referrals)‏ – Sensors: Keyboard (entry of symptoms, findings, patient's answers)‏ 12

13. www.sti-innsbruck.at PEAS – Example 3 Agent: Part-picking robot – Performance measure: Percentage of parts in correct bins – Environment: Conveyor belt with parts, bins – Actuators: Jointed arm and hand – Sensors: Camera, joint angle sensors 13

14. www.sti-innsbruck.at PEAS Agent: Interactive English tutor – Performance measure: Maximize student's score on test – Environment: Set of students – Actuators: Screen display (exercises, suggestions, corrections)‏ – Sensors: Keyboard 14

15. www.sti-innsbruck.at Environment types Fully observable (vs. partially observable): An agent's sensors give it access to the complete state of the environment at each point in time. Deterministic (vs. stochastic): The next state of the environment is completely determined by the current state and the action executed by the agent. (If the environment is deterministic except for the actions of other agents, then the environment is strategic)‏ Episodic (vs. sequential): The agent's experience is divided into atomic "episodes" (each episode consists of the agent perceiving and then performing a single action), and the choice of action in each episode depends only on the episode itself not on the whole history. 15

16. www.sti-innsbruck.at Environment types Static (vs. dynamic): The environment is unchanged while an agent is deliberating. (The environment is semi-dynamic if the environment itself does not change with the passage of time but the agent's performance score does)‏ Discrete (vs. continuous): A limited number of distinct, clearly defined percepts and actions. Single agent (vs. multiagent): An agent operating by itself in an environment. 16

17. www.sti-innsbruck.at Environment types Chess with Chess without Taxi drivinga clock Fully observableYesYesNo DeterministicStrategicStrategicNo Episodic NoNoNo Static SemiYes No DiscreteYes YesNo Single agentNoNoNo The environment type largely determines the agent design The real world is (of course) partially observable, stochastic, sequential, dynamic, continuous, multi-agent 17

18. www.sti-innsbruck.at Agent functions and programs An agent is completely specified by the agent function mapping percept sequences to actions One agent function (or a small equivalence class) is rational Aim: find a way to implement the rational agent function concisely 18

19. www.sti-innsbruck.at Table-lookup agent function Table-Driven-Agent(percept) returns an action static percepts, a sequence, initially empty table, a table of actions, indexed by percept sequences, intially fully specified append percept to the end of percepts action <- Lookup(percepts, table)‏ return action Drawbacks: – Huge table, – Take a long time to build the table, maybe even impossible to build (infinite)‏ – No autonomy – Even with learning, need a long time to learn the table entries 19

20. www.sti-innsbruck.at Agent program for a vacuum-cleaner agent function Reflex-Vacuum-Agent( [location,status]) returns an action if status = Dirty then return Clean else if location = A then return Right else if location = B then return Left 20

21. www.sti-innsbruck.at Agent types Four basic types in order of increasing generality: – Simple reflex agents – Model-based reflex agents – Goal-based agents – Utility-based agents 21

22. www.sti-innsbruck.at Simple reflex agents 22

23. www.sti-innsbruck.at Simple reflex agents function Simple-Reflex-Agent(percept) returns an action static rules, a set of condition-action rules state <- Interpret-Input(percept )‏ rule <- Rule-Match(state,rules)‏ action <- Rule-Action[rule] return action Drawbacks: – It only work in general if the environment is fully observable 23

24. www.sti-innsbruck.at Model-based reflex agents 24

25. www.sti-innsbruck.at Model-based reflex agents function Reflex-Agent-With-State(percept) returns an action static state, a description of the current world state rules, a set of condition-action rules action, the most recent action, initially none state <- Update-State(state, action, percept)‏ rule <- Rule-Match(state,rules)‏ action <- Rule-Action[rule] return action Difference: The agent keeps track of an internal state, i.e. the internal model of the world! 25

26. www.sti-innsbruck.at Goal-based agents 26

27. www.sti-innsbruck.at Utility-based agents 27

28. www.sti-innsbruck.at Learning agents 28

29. www.sti-innsbruck.at Illustration: the Wumpus world Performance measure – gold +1000, death -1000 – -1 per step, -10 for using the arrow Environment – Squares adjacent to wumpus are smelly – Squares adjacent to pit are breezy – Glitter iff gold is in the same square – Shooting kills wumpus if you are facing it – Shooting uses up the only arrow – Grabbing picks up gold if in same square – Releasing drops the gold in same square Sensors: Stench, Breeze, Glitter, Bump, Scream Actuators: Left turn, Right turn, Forward, Grab, Release, Shoot

30. www.sti-innsbruck.at Illustration: the Wumpus world (2) World characterization: Fully Observable No – only local perception Deterministic Yes – outcomes exactly specified Episodic No – sequential at the level of actions Static Yes – Wumpus and Pits do not move Discrete Yes Single-agent? Yes – Wumpus is essentially a natural feature

31. www.sti-innsbruck.at Extensions The following extension can be considered: Schema-based agents have a structured description of the world. Web services can be considered as agents on the Web. Peer-to-peer systems where each peer is an agent. Logical Agents reson on logical facts to deduce information on the environment

32. www.sti-innsbruck.at 32 Conclusions We have learnt –What are agents and what is meant by their environments –What is a rational agent –How to describe an agent in terms of PEAS‏ –The various environment types and agent types

33. www.sti-innsbruck.at 33 References The lecture book J. Ferber, Multi-Agents Systems: an Introduction to Distributed Artificial Intelligence