Paper page - RoboSPA: Can VLA Models Go Beyond Simple Scenes and Short-Horizon Tasks?

Source: https://huggingface.co/papers/2609.05324 Authors:

,

,

,

,

,

,

,

,

,

,

Abstract

RoboSPA is a large-scale robotic manipulation benchmark that evaluates vision-language-action models on fine-grained spatial reasoning and long-horizon procedural planning across progressively harder task variants.

Vision-Language-Action (VLA) models have shown promising progress in language-conditioned robotic manipulation. However, existing datasets and benchmarks mainly evaluate task completion under predefined settings, offering limited insight into model reasoning under increasing spatial and procedural complexity. We introduce RoboSPA (Robot Spatial-Procedural Assessment), a large-scale robotic manipulation dataset and benchmark for diagnosingembodied reasoningin VLA models. RoboSPA focuses on two core dimensions, Fine-GrainedSpatial Reasoningand Long-HorizonProcedural Planning, covering 10 task categories and 56 base tasks. Each task is instantiated across five difficulty levels, yielding 280 variants with increasing spatial ambiguity and procedural complexity. We collect 527K trajectories across multiple embodiments and diverse scenes. Beyond binary success rate, RoboSPA introducesdiagnostic metricsfor more detailed evaluation. Experiments on representative VLA models show that current systems still struggle with complex spatial relations, precise low-level execution, and memory-intensive planning. These results establish RoboSPA as a challenging diagnostic benchmark for developing more capable, reliable, and generalizable embodied agents. Our data and code are available at https://github.com/fanzhenxuan/RoboSPA.

View arXiv pageView PDFProject pageGitHub7Add to collection

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2609.05324 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2609.05324 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2609.05324 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.