Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Multi-threaded Recast-Based A* Pathfinding for Scalable Navigation in Dynamic Game Environments

Tiroshan Madushanka, Sakuna Madushanka

TL;DR

The paper tackles the performance-realism trade-off in A* pathfinding for dynamic 3D game environments. It introduces a multi-threaded framework that combines Recast-based NavMesh graphs, Funnel and Bezier post-processing, and density-aware crowd coordination to sustain high frame rates while producing smooth, collision-free movement. Key contributions include Recast Graph generation, asynchronous multi-threaded A*, Bezier trajectory smoothing with $B(t)$ interpolation, and a density analysis module for crowd management, validated across ten progressive phases that show robust scalability up to 1000 agents and 350+ FPS. The results demonstrate practical impact by enabling scalable, realistic navigation in dynamic worlds, with implications for large-scale crowds and real-time gameplay.

Abstract

While the A* algorithm remains the industry standard for game pathfinding, its integration into dynamic 3D environments faces trade-offs between computational performance and visual realism. This paper proposes a multi-threaded framework that enhances standard A* through Recast-based mesh generation, Bezier-curve trajectory smoothing, and density analysis for crowd coordination. We evaluate our system across ten incremental phases, from 2D mazes to complex multi-level dynamic worlds. Experimental results demonstrate that the framework maintains 350+ FPS with 1000 simultaneous agents and achieves collision-free crowd navigation through density-aware path coordination.

Multi-threaded Recast-Based A* Pathfinding for Scalable Navigation in Dynamic Game Environments

TL;DR

The paper tackles the performance-realism trade-off in A* pathfinding for dynamic 3D game environments. It introduces a multi-threaded framework that combines Recast-based NavMesh graphs, Funnel and Bezier post-processing, and density-aware crowd coordination to sustain high frame rates while producing smooth, collision-free movement. Key contributions include Recast Graph generation, asynchronous multi-threaded A*, Bezier trajectory smoothing with interpolation, and a density analysis module for crowd management, validated across ten progressive phases that show robust scalability up to 1000 agents and 350+ FPS. The results demonstrate practical impact by enabling scalable, realistic navigation in dynamic worlds, with implications for large-scale crowds and real-time gameplay.

Abstract

While the A* algorithm remains the industry standard for game pathfinding, its integration into dynamic 3D environments faces trade-offs between computational performance and visual realism. This paper proposes a multi-threaded framework that enhances standard A* through Recast-based mesh generation, Bezier-curve trajectory smoothing, and density analysis for crowd coordination. We evaluate our system across ten incremental phases, from 2D mazes to complex multi-level dynamic worlds. Experimental results demonstrate that the framework maintains 350+ FPS with 1000 simultaneous agents and achieves collision-free crowd navigation through density-aware path coordination.
Paper Structure (49 sections, 2 equations, 4 figures, 5 tables)

This paper contains 49 sections, 2 equations, 4 figures, 5 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Pathfinding Algorithms
  4. Search Space Representation
  5. Trajectory Smoothing and Crowd Simulation
  6. Methodology
  7. NavMesh-Based Search Space
  8. Dynamic Updates
  9. Recast Graph Generation
  10. A* Implementation with Multi-threading
  11. Thread Pool Architecture
  12. Work Queue Management
  13. Path Post-Processing Pipeline
  14. Start and End Point Modification
  15. Funnel Algorithm
  16. ...and 34 more sections

Figures (4)

  • Figure 1: System architecture showing the processing pipeline from environment input through graph generation, path computation, smoothing, and movement control with feedback loops for dynamic updates.
  • Figure 2: Density analysis scenarios: (Top) Distributed movement with 100 agents navigating to individual destinations, (Middle) Opposing groups of 50 agents crossing and switching sides, (Bottom) Convergence scenario with agents moving toward center point showing final stable distribution.
  • Figure 3: Frame rate comparison between single-threaded baseline and proposed multi-threaded architecture across varying agent counts. The baseline degrades below 20 FPS at 200 agents while the proposed framework maintains above 90 FPS.
  • Figure 4: Frame rate scaling with NPC count. The system maintains playable rates (350 FPS) up to 1000 agents, well above the 60 FPS threshold.