Adaptable Hierarchical Task Network for Dynamic Game Planning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing hierarchical task networks in video games struggle to adapt plans effectively when environmental changes occur, leading to frequent replanning and loss of long-term consistency in character control.
Innovation Solution
A plan processing program that generates plans with adaptable tasks based on goal information, allowing subsequent tasks to change according to environmental states or conditions, using a hierarchical task network and a simulator for logical inference and scenario planning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional hierarchical task networks are used for character control planning, then the system structure is simple and easy to implement, but the plan cannot adapt effectively to environmental changes, causing frequent replanning and loss of long-term consistency
Solution Approach 1:
The patent applies dynamics by making the task network adaptable and reconfigurable during execution. The system dynamically adjusts the hierarchical task structure based on environmental changes, allowing tasks to be modified, added, or removed while maintaining overall plan coherence. This enables the character controller to respond flexibly to changing conditions without complete replanning, preserving long-term consistency while adapting to new situations.
Solution Approach 2:
The patent implements feedback mechanisms that continuously monitor environmental conditions and plan execution status. When environmental changes are detected, the system feeds this information back to the planning layer, which then adjusts the task network accordingly. This closed-loop feedback ensures that the character controller maintains long-term consistency by making incremental adaptations rather than frequent complete replans.
2Adaptability or versatility
If frequent replanning is performed to handle environmental changes, then adaptability improves, but computational overhead increases and character behavior loses consistency
Solution Approach 1:
The patent segments the planning system into hierarchical layers where high-level strategic decisions are made less frequently and low-level tactical adjustments are made more frequently. This segmentation allows the system to handle environmental changes efficiently by making localized adjustments at lower levels without triggering complete replanning at higher levels, reducing computational overhead while maintaining adaptability.
Solution Approach 2:
The system dynamically determines the appropriate level and frequency of replanning based on the significance of environmental changes. Minor changes are handled through local task adjustments, while major changes trigger more comprehensive replanning. This dynamic approach optimizes computational resource usage and maintains character behavior consistency by avoiding unnecessary frequent replanning.
3Productivity
If the plan is generated based on current state only, then the planning process is simple and fast, but the character cannot maintain long-term consistency when environment changes
Solution Approach 1:
The patent applies preliminary action by pre-defining reusable task templates and subplans that can be quickly instantiated and adapted. Instead of generating plans from scratch, the system prepares a library of pre-validated task sequences that can be rapidly deployed and modified based on environmental conditions. This approach maintains fast planning speed while improving adaptability through pre-prepared response patterns.
Solution Approach 2:
The patent implements nesting by organizing tasks into hierarchical levels where higher-level tasks contain lower-level subtasks. When environmental changes occur, the system can adjust at any hierarchical level without affecting the entire plan structure. This nested organization enables fast planning by allowing localized modifications within the hierarchy, maintaining both planning speed and adaptability to environmental changes.
Data Source
AI summary
A plan processing program for planning a task executed by a character based on a hierarchical task network causes a server to implement a plan generation function of generating a plan configured with a plurality of tasks from a domain based on goal information, in which in the plan generation function, a function of generating the plan in which a subsequent task to be employed is configured to change in accordance with a state or a condition is implemented.