2.3 The Properties of Task Environments (BT104CO)

1. Understanding the Task Environment

In the framework of Russell and Norvig, the design of an agent is strictly dictated by the Task Environment. To design a rational agent, we must first categorize the environment using six key dimensions.

01

Fully vs. Partially Observable

Fully: Sensors detect all relevant aspects (e.g., Chess).

Partially: Missing info due to noise or limited range (e.g., Poker).

02

Single Agent vs. Multiagent

Single: Acting alone (e.g., Crossword).

Multiagent: Other agents are present. Can be Competitive (Chess) or Cooperative (Traffic).

03

Deterministic vs. Stochastic

Deterministic: Next state is completely certain (e.g., Sudoku).

Stochastic: Uncertainty or randomness involved (e.g., Taxi driving).

Strategic: Deterministic except for other agents.
04

Episodic vs. Sequential

Episodic: Actions are independent "episodes" (e.g., Image Analysis).

Sequential: Decisions have long-term consequences (e.g., Driving).

05

Static vs. Dynamic

Static: World doesn't change while thinking (e.g., Crossword).

Dynamic: World keeps moving during deliberation (e.g., Taxi).

Semidynamic: Only performance score changes (e.g., Timed Chess).
06

Discrete vs. Continuous

Discrete: Limited, distinct number of states (e.g., Chess).

Continuous: Smoothly varying speed, angles, coords (e.g., Self-driving cars).

2. Summary Table of Task Environments

Environment Observable Agents Deterministic Episodic Static Discrete
Chess Fully Multi Strategic Sequential Static Discrete
Taxi Driving Partial Multi Stochastic Sequential Dynamic Continuous
Medical Diagnosis Partial Single Stochastic Sequential Dynamic Continuous
Image Analysis Fully Single Deterministic Episodic Semi Continuous

3. Why This Matters for AI Design

The Easy Path

"Toy Worlds"

Fully Observable, Deterministic, Static, Discrete.

Simple search algorithms and "Simple Reflex" architectures can solve these perfectly.

The Challenge

The Real World

Partially Observable, Stochastic, Dynamic, Continuous.

Requires sophisticated "Model-Based" or "Utility-Based" agents that handle uncertainty.

Exam Tip

If asked to describe an environment (like a "Vacuum Cleaner"), use these six terms as your technical headings to provide a complete answer.

Practice Quiz