Development of the 3D Game “Lavender's Warmth” Using the Collision Detection Method

Abstract

This study presents the development of a 3D puzzle–adventure game titled “Lavender’s Warmth” using the Collision Detection and Finite-State Machine (FSM) methods. The use of Collision Detection is essential because the game relies heavely in physical interaction between puzzle pieces, slots, and environmental object. Without Collision Detection, the game would fail to validate puzzle placement. Meanwhile, the FSM approach is required to regulate enemy behaviour in structured manner. The Finite-State Machine was chosen because it is one of the most widely adopted approaches for modeling NPC behavior, offering deterministic transitions, low memory usage, and ease of debugging. Alternative techniques such as behavior trees or utility AI are more complex and unnecessary for the simple enemy mechanics in this game. Therefore, Finite-State Machine provides the most appropriate balance between functionality, performance, and development simplicity. The game was developed using Unity 3D and tested through functionality, method, and user evaluations. The results showed that all main features worked as expected, with 52.63% of users strongly agreeing and 40.64% agreeing that the game was engaging and enjoyable. The implementation of both methods successfully enhanced interactivity, responsiveness, and gameplay consistency.