Unreal Engine 4 AI Programming Essentials

The book emphasizes using a blackboard for AI data storage and retrieval; however in larger games with multiple AI agents and shared or partially overlapping knowledge, how does one balance per agent blackboard data with a global “world state” representation? And also are there architectural patterns or data structures recommended to optimize concurrency and avoid data duplication? The text doesn’t really dwelve into customizing or extending the AI Perception sensing models for more nuanced AI behavior such as partial inclusion or dynamic lighting conditions and thereby, doesn’t really address how this might be integrated without substantially affecting performance. I agree that environment query systems (EQS) is powerful for dynamic environment queries, but multi-criteria path finding or real time combat positioning can be performance intensive so itd be helpful to know how to handle optimization strategies such as partial carriers or hierarchical partitioning and see if there’s any recommended design patterns for keeping EQS queries maintainable as complexity grows.