Skip to content

Open-Source

IILABS 3D: iilab Indoor LiDAR-based SLAM Dataset

Indoor environments present unique challenges for Simultaneous Localization and Mapping (SLAM). Existing datasets typically focus on outdoor scenarios and rely on a single LiDAR sensor, limiting the evaluation of SLAM algorithms in complex indoor settings. The IILABS 3D dataset was designed to address these limitations by providing a comprehensive, sensor-rich dataset for benchmarking 3D LiDAR-based SLAM algorithms indoors. It includes data from four different 3D LiDAR sensors (Velodyne VLP-16, Ouster OS1-64, RoboSense RS-Helios-5515, and Livox Mid-360), along with an IMU and wheel odometry, all recorded using a wheeled mobile robot in the iilab. High-precision ground-truth poses were acquired using a OptiTrack Motion Capture system. The dataset also contains calibration sequences and six benchmark trajectories, enabling reproducible and rigorous SLAM evaluations. To support the use of this dataset, a complete open-source toolkit was developed, providing scripts for data handling, metric computation (ATE, RTE, RRE), and SLAM algorithm benchmarking. Additionally, a benchmark suite was implemented to evaluate nine state-of-the-art SLAM algorithms using the dataset.

Links:

Citation (Dataset):

Ribeiro, J.D., Sousa, R.B., Martins, J.G., Aguiar, A.S., Santos, F.N., & Sobreira, H.M. (2025). IILABS 3D: iilab Indoor LiDAR-based SLAM Dataset. DOI: 10.25747/VHNJ-WM80.