Patrol Paths
Patrol Paths in Ninja Bot are defined using Patrol Point Actors (ANinjaBotPatrolPointActor) placed directly in the level. Each patrol point can reference a subsequent patrol point, forming a simple directed path. A patrol point may also terminate the path by not referencing a successor.
In addition to defining movement flow, patrol points can optionally inject additional bot logic by providing dynamic Behavior Trees or State Trees. This allows designers to alter or extend behavior contextually as a bot reaches specific points along a patrol route.
Patrol Point Actors support Gameplay Tags, enabling bots to discover compatible patrol paths at runtime. Agents can use EQS queries to search for patrol points of the appropriate type within a given range and filter them by tags, allowing patrol behavior to adapt dynamically to the environment or agent role.
Patrol Point Properties
Patrol Point Actors expose a small set of properties that define how they participate in patrol paths and optionally influence bot behavior.
Property | Description |
|---|---|
Patrol Point Tags | Gameplay tags used to identify and categorize this patrol point. These tags can be used by EQS queries or other discovery logic to find compatible patrol paths. |
Logic Type | Defines whether this patrol point provides additional logic to the bot. When set to |
Behavior Tree Asset | Behavior Tree provided by this patrol point when the logic type is set to Behavior Tree. The tree can be dynamically injected by compatible bots. |
State Tree Asset | State Tree provided by this patrol point when the logic type is set to State Tree. The tree can be dynamically injected by compatible bots. |
Next Patrol Point | Reference to the next patrol point in the path. When unset, the patrol route terminates at this point. |
Patrol Points support the following logic types:
Logic Type | Description |
|---|---|
None | The patrol point does not provide any additional logic and acts only as a navigation target. |
Behavior Tree | Provides a dynamic Behavior Tree to be injected when the bot reaches this patrol point. Compatible only with bots using Behavior Trees as their brain. |
State Tree | Provides a dynamic State Tree to be injected when the bot reaches this patrol point. Compatible with bots using either Behavior Trees or State Trees. |
AI Integration
Ninja Bot provides a set of Behavior Tree nodes that allow patrol logic to be driven directly from Patrol Point Actors. These nodes focus on dynamic behavior injection and patrol progression, keeping patrol paths data-driven.
The Set Patrol Point Behavior Service attempts to inject a dynamic Behavior Tree provided by the current patrol point. When the service becomes relevant:
It reads the patrol point from the Blackboard, usually obtained via an Environment Query.
If the patrol point provides a dynamic Behavior Tree, it injects it using the configured tag.
If no tree is provided by the actor, it attempts to inject one from the bot setup.
If no tree is available from the setup, it will use the default value set in the task.
Property | Description |
|---|---|
Dynamic Injection Tag | Gameplay tag used to inject the dynamic behavior. Must match the injection point defined in the Behavior Tree. |
Patrol Point Key | Blackboard key referencing the patrol point actor providing the behavior. |
The Advance Patrol Point Task progresses the patrol route by resolving the next patrol point from the current one. It updates the patrol point reference in the Blackboard, allowing the bot to continue along a patrol as the branch restarts.
Property | Description |
|---|---|
Patrol Point Key | Blackboard key referencing the current patrol point. This value is updated to the next patrol point when the task executes. |
Ninja Bot provides a set of State Tree tasks that enable patrol logic to be driven directly by Patrol Point Actors. These tasks allow patrol paths to **inject contextual logic** and advance patrol progression in a data-driven manner.
The Set Patrol Point Behavior Task attempts to inject a dynamic State Tree provided by the current patrol point. When the state is entered, the following sequence occurs:
It reads the patrol point from the context, usually obtained through an Environment Query.
If the patrol point provides a dynamic State Tree, it injects it using the configured Dynamic Injection Tag.
If no State Tree is provided by the patrol point actor, it attempts to inject one from the Bot Setup.
If no State Tree is available from the Bot Setup, it falls back to the default value configured on the state.
Property | Description |
|---|---|
Patrol Point | Input. Reference to the actor representing the current patrol point. |
Dynamic Injection Tag | Gameplay tag used to inject the dynamic State Tree. Must match the State Tree node configured as a linked asset. |
The Advance Patrol Point Task progresses the patrol route by resolving the next patrol point from the current one. It updates the patrol point reference in the context, allowing the Bot to continue along the patrol route as the state restarts.
Property | Description |
|---|---|
Patrol Point | Input. Reference to the current patrol point actor. |
Result | Output reference where the next patrol point actor is written. |
Smart Objects Integration
Patrol Points are implemented as lightweight target point actors rather than Smart Objects. This design choice prioritizes simplicity and avoids the complexity of slot ownership and reservation during early design. If a project requires Smart Object integration or Gameplay Behaviors, Patrol Points can be extended by attaching the appropriate Smart Object components and interacting with them as usual.