SanD-Planner: Sample-Efficient Diffusion Planner in B-Spline Space for Robust Local Navigation

TL;DR

SanD-Planner is a sample-efficient diffusion-based local planner that uses B-spline space for smooth, reliable navigation in cluttered environments, requiring minimal training data and demonstrating strong transferability.

Contribution

It introduces SanD-Planner, a novel diffusion-based local planning method operating in B-spline space, significantly reducing training data requirements and improving robustness.

Findings

Achieves 90.1% success in simulated cluttered environments.

Attains 72.0% success in indoor simulations.

Requires only 0.25% of demonstration data compared to baseline.

Abstract

The challenge of generating reliable local plans has long hindered practical applications in highly cluttered and dynamic environments. Key fundamental bottlenecks include acquiring large-scale expert demonstrations across diverse scenes and improving learning efficiency with limited data. This paper proposes SanD-Planner, a sample-efficient diffusion-based local planner that conducts depth image-based imitation learning within the clamped B-spline space. By operating within this compact space, the proposed algorithm inherently yields smooth outputs with bounded prediction errors over local supports, naturally aligning with receding-horizon execution. Integration of an ESDF-based safety checker with explicit clearance and time-to-completion metrics further reduces the training burden associated with value-function learning for feasibility assessment. Experiments show that training with $500$ episodes (merely $0.25\%$ of the demonstration scale used by the baseline), SanD-Planner achieves state-of-the-art performance on the evaluated open benchmark, attaining success rates of $90.1\%$ in simulated cluttered environments and $72.0\%$ in indoor simulations. The performance is further proven by demonstrating zero-shot transferability to realistic experimentation in both 2D and 3D scenes. The dataset and pre-trained models will also be open-sourced.