AI - Behavior Design Prototype
Project Overview (Prototyping Phase)
This showcase focuses on the **Behavioral Design** and **Decision Logic** of the Minotaur agent during a prototyping phase. For this specific iteration, the **Panda AI Engine** was utilized to rapidly map out complex sensory-response loops before migrating the logic into a more specialized internal architecture.
Key Features & Logic:
- Reactive Sound Detection: The agent continuously monitors for environmental sound events. When a sound is detected, it triggers an investigation state, guiding the agent to the sound origin.
- Patrol-to-Alert Lifecycle: A hierarchical tree that manages transitions between low-alert patrolling and high-alert direct player engagement.
- Steering & Navigation: Integration of sensory data with navigation systems to ensure movement feels deliberate and weighted for a large creature.
Architecture Evolution
Note for Technical Reviewers: While this prototype leveraged the Panda middleware for rapid behavioral mapping, **current versions of the game have evolved to use a fully custom Utility AI (Utility Behavior) system.** This migration allowed for more granular weighting of actions based on a wider range of world variables, moving away from the rigid structure of the initial prototype.