COMP9414/ 9814/ 3411: Artificial Intelligence. 2. Environment Types. UNSW c Alan Blair,

Transcription

1 COMP9414/ 9814/ 3411: rtificial Intelligence 2. Environment Types

2 COMP9414/9814/ s1 Environments 1 gent Model sensors environment percepts actions? agent actuators

3 COMP9414/9814/ s1 Environments 2 The PES model of an gent Performance measure Environment ctuators Sensors

4 COMP9414/9814/ s1 Environments 3 gents as functions gents can be evaluated empirically, sometimes analysed mathematically gent is a function from percept sequences to actions Ideal rational agent would pick actions which are expected to maximise the performance measure.

5 COMP9414/9814/ s1 Environments 4 Example: utomated Taxi Performance measure: safety, reach destination, maximize profits, obey laws, passenger comfort,... Environment: city streets, freeways, traffic, pedestrians, weather, customers,... ctuators: steer, accelerate, brake, horn, speak/display,... Sensors: video, accelerometers, gauges, engine sensors, keyboard, GPS,...

6 COMP9414/9814/ s1 Environments 5 Examples of I Tasks Path Search Problems Games Wumpus World Robots Web gents

7 COMP9414/9814/ s1 Environments 6 Path Search Problems Oradea Neamt Zerind rad Iasi Sibiu Fagaras Vaslui Timisoara Lugoj Rimnicu Vilcea Pitesti Start Goal Mehadia Urziceni Hirsova Dobreta Bucharest Craiova Giurgiu Eforie

8 COMP9414/9814/ s1 Environments 7 Path Search in Configuration Space Start State Goal State

9 COMP9414/9814/ s1 Environments 8 Constraint Satisfaction Problems

10 COMP9414/9814/ s1 Environments 9 Games Chess Dice games Card games

11 COMP9414/9814/ s1 Environments 10 Environment types We can classify environments as: - Fully Observable vs. Partially Observable - Deterministic vs. Stochastic - Single-gent vs. Multi-gent - Episodic vs. Sequential - Static vs. Dynamic - Discrete vs. Continuous - Known vs. Unknown - Simulated vs. Situated or Embodied

12 COMP9414/9814/ s1 Environments 11 Environment types Fully Observable: percept contains all relevant information about the world Deterministic: current state of world uniquely determines the next Episodic: only the current (or recent) percept is relevant Static: environment doesn t change while the agent is deliberating Discrete: finite (or countable) number of possible percepts/actions Known: the rules of the game, or physics/dynamics of the environment are known to the agent Simulated: a separate program is used to simulate an environment, feed percepts to agents, evaluate performance, etc.

13 COMP9414/9814/ s1 Environments 12 Environment types Fully Observable Deterministic Multi-gent Episodic Static Discrete Known Simulated Rubik s Chess Dice Card Wumpus Robocup Stock Cube Games Games World Soccer Trading The real world is (of course) partially observable, stochastic, multi-agent sequential, dynamic, continuous, unknown, situated and embodied.

14 COMP9414/9814/ s1 Environments 13 Example I Environment - Wumpus World Environment Squares adjacent to Wumpus are Smelly Squares adjacent to Pit are Breezy Glitter iff Gold is in the same square Shoot kills Wumpus if you are facing it uses up the only arrow Grab picks up Gold if in same square Stench Breeze Stench Gold Stench Breeze STRT Breeze PIT Breeze PIT PIT Breeze Breeze

15 COMP9414/9814/ s1 Environments 14 Wumpus World PES description Performance measure Return with Gold +1000, death per step, -10 for using the arrow ctuators Left, Right, Forward, Grab, Shoot Sensors Breeze, Glitter, Stench

16 COMP9414/9814/ s1 Environments 15 Exploring a Wumpus World

17 COMP9414/9814/ s1 Environments 16 Exploring a Wumpus World B

18 COMP9414/9814/ s1 Environments 17 Exploring a Wumpus World P? B P?

19 COMP9414/9814/ s1 Environments 18 Exploring a Wumpus World P? B P? S

20 COMP9414/9814/ s1 Environments 19 Exploring a Wumpus World P P? B P? S W

21 COMP9414/9814/ s1 Environments 20 Exploring a Wumpus World P P? B P? S W

22 COMP9414/9814/ s1 Environments 21 Exploring a Wumpus World P P? B P? S W

23 COMP9414/9814/ s1 Environments 22 Exploring a Wumpus World P P? B P? BGS S W

24 COMP9414/9814/ s1 Environments 23 Robots

25 COMP9414/9814/ s1 Environments 24 Situated and Embodied Cognition Rodney Brooks 1991: Situatedness: The robots are situated in the world they do not deal with abstract descriptions, but with the here and now of the environment which directly influences the behaviour of the system. Embodiment: The robots have bodies and experience the world directly their actions are part of a dynamics with the world, and actions have immediate feedback on the robot s own sensations.

26 COMP9414/9814/ s1 Environments 25 Situated vs. Embodied Situated but not Embodied: High frequency stock trading system: it deals with thousands of buy/sell bids per second and its responses vary as its database changes. but it interacts with the world only through sending and receiving messages. Embodied but not Situated: an industrial spray painting robot: does not perceive any aspects of the shape of an object presented to it for painting; simply goes through a pre-programmed series of actions but it has physical extent and its servo routines must correct for its interactions with gravity and noise present in the system.

27 COMP9414/9814/ s1 Environments 26 State of the art Which of the following can be done at present? Play a decent game of table tennis (ping-pong) Drive in the center of Cairo, Egypt Drive along a curving mountain road Play games like Chess, Go, Bridge, Poker Discover and prove a new mathematical theorem Write an intentionally funny story Give competent legal advice in a specialized area of law Translate spoken English into spoken Swedish (or Chinese) in real time

28 COMP9414/9814/ s1 Environments 27 Summary Environments can be classified in terms of whether they are observable, deterministic, single- or multi- agent, episodic, static, discrete, known, simulated. The environment type strongly influences the agent design (discussed in the next lecture..)