This paper presents a method for evaluating the Point-LIO algorithm on a Mecanum-wheeled mobile robot, based on results obtained from both manual mapping and autonomous mapping experiments. The testing environment was deliberately designed to be spatially complex, featuring challenging structures such as an open atrium, metal railings, narrow corridors, and a semi-indoor/semi-outdoor layout that introduces both geometric and environmental variability. The results demonstrate that Point-LIO delivers robust and accurate localization, with mean absolute positioning errors consistently remaining below 7 cm. The mapping accuracy also remained stable across various test regions. When integrated with the OctoMap framework, the system successfully enabled autonomous navigation and full 3D map generation, showing that the algorithm can support real-time SLAM in real-world settings. However, the experiments also revealed a gradual drift along the Z-axis in certain areas farther from the origin, along with a slight decrease in localization accuracy as the robot moved away from its initial reference point. These findings suggest that while Point-LIO is highly effective, future improvements may focus on mitigating Z-axis drift and long-range localization degradation.