UA-MPC: Uncertainty-Aware Model Predictive Control for Motorized LiDAR Odometry

TL;DR

This work tackles the challenge of achieving accurate 3D sensing with motorized LiDAR by introducing UA-MPC, an uncertainty-aware model predictive control framework that adapts motor speed based on predicted LiDAR odometry uncertainty across viewing orientations. It predicts $U(\theta_i)$ via ray-traced panoramic depth maps from the local LO map and optimizes a horizon-based objective $F(\mathbf{\Omega}_j)$ using a surrogate $U'$ to enable fast edge-computing control. A ROS-based realistic simulation environment built on MARSIM and the MCD dataset enables robust, scalable evaluation of motor-control strategies. Experiments on both simulated scenes and real handheld hardware show UA-MPC achieving over 60% lower ATE with less than 2% reduction in scanning efficiency compared to constant-speed control, and demonstrate robustness in challenging, feature-sparse environments.

Abstract

Accurate and comprehensive 3D sensing using LiDAR systems is crucial for various applications in photogrammetry and robotics, including facility inspection, Building Information Modeling (BIM), and robot navigation. Motorized LiDAR systems can expand the Field of View (FoV) without adding multiple scanners, but existing motorized LiDAR systems often rely on constant-speed motor control, leading to suboptimal performance in complex environments. To address this, we propose UA-MPC, an uncertainty-aware motor control strategy that balances scanning accuracy and efficiency. By predicting discrete observabilities of LiDAR Odometry (LO) through ray tracing and modeling their distribution with a surrogate function, UA-MPC efficiently optimizes motor speed control according to different scenes. Additionally, we develop a ROS-based realistic simulation environment for motorized LiDAR systems, enabling the evaluation of control strategies across diverse scenarios. Extensive experiments, conducted on both simulated and real-world scenarios, demonstrate that our method significantly improves odometry accuracy while preserving the scanning efficiency of motorized LiDAR systems. Specifically, it achieves over a 60\% reduction in positioning error with less than a 2\% decrease in efficiency compared to constant-speed control, offering a smarter and more effective solution for active 3D sensing tasks. The simulation environment for control motorized LiDAR is open-sourced at: \url{https://github.com/kafeiyin00/UA-MPC.git}.

Figures (9)

• Figure 1: Coordinates and mechanical design of the proposed motorized LiDAR system.
• Figure 2: Workflow of Uncertainty-Aware Model Predictive Control (UA-MPC).
• Figure 3: Prediction of the uncertainty of the LiDAR Odometry (LO) at different observing orientations controlled by the motor. (a) Taking the local map from LO as the input. (b) Rendering the panoramic depth map using the local map. (c) Sampling of the LiDAR measurements at specific orientation to calculate $U(\theta_j)$. (d) The typical laser scanner with different vertical and horizontal Field of View (FoV).
• Figure 4: Surrogate function $U^\prime$, a piecewise linear function, is used for approximating the raw uncertainty function $U$.
• Figure 5: Simulation environment for motorized LiDAR system based on MARSIM and ROS.

Theorems & Definitions (2)

• Remark 1
• Remark 2