A Collision-Free Sway Damping Model Predictive Controller for Safe and Reactive Forestry Crane Navigation

TL;DR

The paper addresses safe forestry crane navigation by jointly damping payload sway and avoiding collisions in cluttered outdoor environments. A collision-free sway-damping MPC is proposed that embeds LiDAR-derived EDF constraints into the optimization, using $ \mathrm{sd}_{i}(\mathbf{q}) > 0 $ with $ \mathrm{sd}_{i}(\mathbf{q}) = \min_{j} \{ d_{\mathrm{EDF}}(\mathbf{p}_{ij}) - r_{ij} \} $ and a pump-flow limit $ Q_{\mathrm{max}} $. The pipeline comprises perception, global planning, and a local MPC, and is validated on a real forestry crane with 10 Hz control. Results show effective sway damping, robust obstacle avoidance, and safe stopping when bypass is not feasible, enabling reactive, environment-aware operation in cluttered outdoor forests.

Abstract

Forestry cranes operate in dynamic, unstructured outdoor environments where simultaneous collision avoidance and payload sway control are critical for safe navigation. Existing approaches address these challenges separately, either focusing on sway damping with predefined collision-free paths or performing collision avoidance only at the global planning level. We present the first collision-free, sway-damping model predictive controller (MPC) for a forestry crane that unifies both objectives in a single control framework. Our approach integrates LiDAR-based environment mapping directly into the MPC using online Euclidean distance fields (EDF), enabling real-time environmental adaptation. The controller simultaneously enforces collision constraints while damping payload sway, allowing it to (i) replan upon quasi-static environmental changes, (ii) maintain collision-free operation under disturbances, and (iii) provide safe stopping when no bypass exists. Experimental validation on a real forestry crane demonstrates effective sway damping and successful obstacle avoidance. A video can be found at https://youtu.be/tEXDoeLLTxA.

Figures (9)

• Figure 1: Forestry crane equipped with LiDAR-based perception system. The proposed MPC framework enables autonomous navigation with simultaneous collision avoidance and sway damping capabilities for safe operation.
• Figure 2: Schematic of our navigation pipeline of the forestry crane. We use the global planner from ecker:2025 together with the EDF-based collision detection routine ecker:iros:2025. To obtain a map, we apply a point cloud filter to remove the crane points, and use OctoMap hornung:2013 in combination with the FIESTA EDF mapping algorithm han:2019. The perception-based collision-free sway damping MPC is the contribution of this work.
• Figure 3: Comparison of the closed-loop behavior using first and second order actuator models. Blue: Measured closed-loop joint velocity. Red dashed: Controller output $u$. Top: Slewing velocity with second order model. Bottom: Slewing velocity with first order model.
• Figure 4: Motions of the passive joint $q_6$ under external force disturbance. Left: MPC is active. Right: MPC is inactive. Red: Motions after single pull. Blue: Motions after triple pull.
• Figure 5: Setup for the sway damping experiment.