Efficient pedestrian evacuation during emergencies involving obstacles, exits, and fire dynamics is critical for safety. Traditional models like Social Force (SF) and Cellular Automaton (CA) have limitations in scalability and decision-making. This study presents a fine-grid CA model in Unity to resolve conflicts and simulate interactions between agents, obstacles, exits, and fire spread. The model’s scoring system minimizes conflicts near exits and explores how exit placement impacts evacuation efficiency, revealing that more exits reduce congestion and evacuation times. This framework offers valuable insights for improving building designs and emergency protocols.