analiza aiqbench
This is a robotics control paper, but its two core mechanisms are learned: an RL-based execution layer that estimates caster states under partial observability, and an uncertainty-aware NMPC that adjusts its costs based on predictive uncertainty. The reported gains come from simulation (Isaac Sim) and limited real-robot validation, so readers should check how well the RL component generalizes beyond the tested caster configurations and contact conditions. The most transferable idea is using learned uncertainty to modulate a model-based controller, which applies to other planning-execution gaps in mobile robotics.
MSC: Multi-Stage Caster-aware Control Framework with Uncertainty-Aware NMPC and RL-based Execution Layer
arXiv:2610.16199v1 Announce Type: cross Abstract: Freely swiveling casters are widely adopted in industrial AMRs/SDVs for their mechanical simplicity and load capacity. However, caster misalignment during low-speed or high-curvature maneuvers induces strong contact nonlinearities, widening the planning-execution gap between commanded $(v, \omega)$ and realized motion. Existing caster-aware strategies are highly sensitive to state-estimation reliability, causing unstable execution in uncertain transition regimes. To address this, we present Multi-Stage Caster-aware Control (MSC), a planning-execution integrated hierarchical framework featuring two core mechanisms. First, an uncertainty-aware NMPC (CU-MPC) dynamically schedules caster-aware costs and input regularization using predictive uncertainty to suppress overly aggressive maneuvers. Second, a reinforcement learning-based execution layer estimates and predicts caster states under partial observability, transforming high-level velocity commands into physically executable wheel speeds adapted to current contact conditions. Validations in Isaac Sim and real robots demonstrate that MSC offers a clear comparative advantage over existing decoupled compensation approaches by significantly improving caster-state estimation accuracy, minimizing misalignment, and enhancing tracking robustness in caster-sensitive maneuvers.
Abstract:Freely swiveling casters are widely adopted in industrial AMRs/SDVs for their mechanical simplicity and load capacity. However, caster misalignment during low-speed or high-curvature maneuvers induces strong contact nonlinearities, widening the planning-execution gap between commanded $(v, \omega)$ and realized motion. Existing caster-aware strategies are highly sensitive to state-estimation reliability, causing unstable execution in uncertain transition regimes. To address this, we present Multi-Stage Caster-aware Control (MSC), a planning-execution integrated hierarchical framework featuring two core mechanisms. First, an uncertainty-aware NMPC (CU-MPC) dynamically schedules caster-aware costs and input regularization using predictive uncertainty to suppress overly aggressive maneuvers. Second, a reinforcement learning-based execution layer estimates and predicts caster states under partial observability, transforming high-level velocity commands into physically executable wheel speeds adapted to current contact conditions. Validations in Isaac Sim and real robots demonstrate that MSC offers a clear comparative advantage over existing decoupled compensation approaches by significantly improving caster-state estimation accuracy, minimizing misalignment, and enhancing tracking robustness in caster-sensitive maneuvers.
Robotics (cs.RO); Machine Learning (cs.LG)
arXiv:2610.16199 [cs.RO]
(or arXiv:2610.16199v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2610.16199 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 252 cuvinte
Subiecte conexe
Unified Planning-Learning Framework for Robust UUV Navigation Under Partial Observability
arXiv · 7 aug. 2026
Digital States 2026: AI Moves From Experiment to Infrastructure
govtech.com · 24 sept. 2026
MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
arXiv · 25 aug. 2026
MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
arXiv · 3 iun. 2026