This paper presents SE(3)-LIVOM, the overall winner of the COMFORT Challenge at IROS 2026. The challenge asks for the trajectory of a legged robot in the GrandTour dataset, whose scenes include collapsed buildings, construction sites and snowfields. SE(3)-LIVOM runs in real time on two LiDARs, one camera and an IMU. Its frontend extends SE(3)-LIO to multiple LiDARs and a camera. Adaptive downsampling keeps the time per scan near 35 ms within and across scenes. An online backend corrects drift with a plane bundle adjustment over submaps and an incremental pose graph. With a single configuration for all scenes, the system reaches an average ATE of 1.40 cm on the six test scenes. The code is available at https://github.com/url-kaist/se3-livom-comfort.
Figures & tables
Split
Validation
Test
Scene
ARC-3
ARC-6
CON-3
EIG-2
SNOW-3
SPX-1
Avg.
ARC-2
ARC-7
CON-4
EIG-1
SNOW-2
SPX-2
Avg.
Sub. ID
FAST-LIVO2 [ 2 ]
1.34
1.24
0.97
2.59
3.44
1.64
1.87
1.22
1.38
3.13
3.32
1.61
2.28
2.16
880291
Ours (w/o backend)
1.24
0.89
1.00
2.10
1.43
1.42
1.35
0.86
1.19
2.30
2.21
1.41
1.99
1.66
933011
Ours
1.20
0.91
0.80
1.70
1.35
1.25
1.20
0.77
1.18
1.68
1.54
1.42
1.81
1.40
935484
Table 1: ATE RMSE (cm) on the validation and test scenes of the GrandTour dataset. Each validation scene was recorded at the site of the test scene in the same column. Ours (w/o backend) is the frontend alone, and Sub. ID is the Codabench submission. Bold is the best and underline the second best.
Figure 2: Frontend time per scan (boxes, left axis) and ATE (markers, right axis) on the validation scenes with two fixed downsampling resolutions and our adaptive one. Boxes span the interquartile range and the lines the 5th to 95th percentiles. The dashed line is the scan period.
Scene
ARC-2
ARC-7
CON-4
EIG-1
SNOW-2
SPX-2
Frontend, mean (ms)
30.1
31.8
42.3
31.6
33.5
27.8
Scans over 100 ms
0
0
10
0
0
0
Backend, longest round (s)
1.4
1.4
19.4
4.4
3.1
0.7
Peak memory (GB)
9.0
7.5
16.7
11.0
6.2
7.9
Table 2: Runtime on the test scenes, bags played at real time on 8 cores and 16 GB. A round is one bundle adjustment, and 100 ms is the scan period.
Real-time LiDAR-visual-inertial odometry and mapping is crucial for navigation and planning tasks in intelligent transportation systems. This study presents a pose-only bundle adjustment (PA) LiDAR-visual-inertial odometry (LVIO), named PA-LVIO, to meet the urgent need for real-time navigation and mapping. The proposed PA framework for LiDAR and visual measurements is highly accurate and efficient, and it can derive reliable frame-to-frame constraints within multiple frames. A marginalization-free and frame-to-map (F2M) LiDAR measurement model is integrated into the state estimator to eliminate odometry drifts. Meanwhile, an IMU-centric online spatial-temporal calibration is employed to obtain a pixel-wise LiDAR-camera alignment. With accurate estimated odometry and extrinsics, a high-quality and RGB-rendered point-cloud map can be built. Comprehensive experiments are conducted on both public and private datasets collected by wheeled robot, unmanned aerial vehicle (UAV), and handheld devices with 28 sequences and more than 50 km trajectories. Sufficient results demonstrate that the proposed PA-LVIO yields superior or comparable performance to state-of-the-art LVIO methods, in terms of the odometry accuracy and mapping quality. Besides, PA-LVIO can run in real-time on both the desktop PC and the onboard ARM computer. The codes and datasets are open sourced on GitHub (https://github.com/i2Nav-WHU/PA-LVIO) to benefit the community.
Hailiang Tang, Tisheng Zhang, Liqiang Wang +3
the GNSS Research Center, Wuhan University, Wuhan 430079, China · the Hubei Technology Innovation Center for Spatiotemporal Information and Positioning Navigation, Wuhan 430079, China · the Hubei Luojia Laboratory, Wuhan 430079, China
This short report documents our entry to the COMFORT Localization Benchmark (IROS 2026), evaluated on the GrandTour dataset recorded with the Boxi payload. It extends MOLA-LO into a LiDAR-inertial system that also ingests IMU and, optionally, legged kinematic odometry. We describe the architecture, the streams consumed, the local protocol that selected the submitted configuration, and the measurements backing our real-time claim.
Jose Luis Blanco-Claraco
Engineering Department University of Almer´ıa, Spain
LiDAR-Inertial-Visual Odometry (LIVO) integrates complementary LiDAR, camera, and IMU measurements to achieve robust state estimation. However, most existing methods assume a fixed relative pose and accurate time synchronization between sensors. Toward modular and heterogeneous robotic systems, this paper presents FREE-LIVO, a tightly coupled odometry method that requires neither a rigid LiDAR-camera rig nor LiDAR-camera time synchronization. The proposed method uses a pretrained image feature matcher to establish correspondences between a camera image and a dense LiDAR intensity image generated from a single LiDAR scan using a generative model. From the resulting 3D point-to-pixel correspondences, we construct reprojection factors that directly constrain LiDAR and camera poses at different acquisition times. These factors are jointly optimized in a single factor graph together with constraints from point cloud registration, visual feature tracking, and the IMU measurements associated with each sensor. On six sequences from the public M3DGR dataset, FREE-LIVO achieves an average absolute trajectory error (ATE) RMSE of 0.45 m without provided LiDAR-camera extrinsic parameters and exhibits minimal accuracy degradation under time synchronization errors of up to 0.5 s. Furthermore, experiments with the LiDAR and camera mounted on separate robots demonstrate that the method can jointly estimate the trajectories of independently moving sensors.
Reon Tabata, Kenji Koide, Shuji Oishi +5
Active Intelligent Systems Lab., Toyohashi University of Technology, Japan · Dept. of Information Technology and Human Factors, AIST, Japan