In our environment, developers primarily have two ways to control agents' movements. Firstly, they can utilize the "set_move" function for step-by-step control of agents, which doesn't rely on additional plugins but requires developers to autonomously manage movement strategies through external code. Secondly, developers can construct a navigation system using the built-in navigation mesh in the UE engine to enable the autonomous movement of agents. Our environment integrates these two control methods. By combining BP function calls for step-by-step control with navigation system control, we can effectively utilize the navigation system to control secondary agents while employing step-by-step control for primary controlled agents. This combined approach offers advantages such as improved environment frame rates in multi-agent scenarios and higher precision control over primary agents.
BP_functions with Navigation system
| BP_function | input | output | help |
|---|---|---|---|
nav_to_goal |
{x} {y} {z} |
Assign the agent a navigation goal, and use Navmesh and AI-controller built-in plugin to automatically control its movement to reach the goal via the shortest path. (The goal has to be reachable in the environment.) | |
nav_to_target |
{target_object} {distance} |
Assign the current agent to a target_object, utilize Navmesh and AI controller to dynamically follow the target_object, stopping at a specified distance from it | |
nav_random |
control agent randomly wandering around | ||
generate_nav_goal |
{ | ||
| ”nav_goal”: “X=0.00 Y=0.00 Z=0.00” | |||
| } | generate a reachable goal location around the agent, return the location. |
Example：
The ground is covered by Nav mesh(Green area), we further utilize the navmesh modifier to segment the area into different colors(only in the scene with clear roads): The agents will prioritize using the pink area for pedestrian navigation tasks, while the blue area will be prioritized for driving navigation tasks.