L6: Decision Making and Behavior Planning

Bridge: From Individual Components to Intelligent Behavior

Learning Objectives

• Design decision-making frameworks under uncertainty

• Implement behavioral planning strategies

• Handle multi-agent interactions

• Ensure safety in critical scenarios

Outline

Part 1: Decision Making Under Uncertainty

Introduces probabilistic and logic-based decision frameworks including Markov Decision Processes (MDPs), Partially Observable MDPs (POMDPs), and Bayesian networks.

Part 2: Behavioral Planning Architectures

Discusses rule-based, state-machine, and hierarchical behavior planners. Covers decision layering and interaction between planning and prediction modules.

Part 3: Multi-Agent Reasoning

Explores coordination among multiple vehicles using game theory, cooperative and non-cooperative strategies, and intent inference models.

Part 4: Safety-Critical Decision Making

Focuses on risk assessment, emergency maneuvers, and formal safety verification methods. Emphasizes compliance with safety envelopes and fail-safe transitions.

Part 5: Real-World Deployment Challenges

Examines challenges in uncertain, unstructured environments such as occlusions, adversarial behaviors, and human unpredictability. Highlights robustness and scalability concerns.