ROBOTNESS
Research

Papers

New robotics and physical-AI papers, with what each one means for the industry.

3 papers
TopicAllManipulation · 20VLA · 17Navigation · 15Perception · 9Robotics · 9Learning · 6Humanoid · 5Safety · 3Simulation · 3Aerial Manipulation · 1Co-speech Humanoid · 1Distributed Manipulation · 1Floorplan SLAM · 1Grounding · 1Humanoid control · 1Locomotion · 1Memory-centric VLA · 1Model-based RL · 1Multi-UAV State Estimation · 1Policy · 1Soft Pneumatic Actuators · 1Tactile · 1Terrain traversability · 1Text-to-3D Policy · 1UAV Navigation · 1UAV traffic monitoring · 1Visual Grounding · 1World models · 1
LevelAnyBeginnerIntermediateExpert
arXiv
SafetyExpert

Neuro-Symbolic Predicate Learning for Semantic Safe Robot Control

Zihan Ye, Jiayi Liu, Puze Liu, Jiayun Li, Georgia Chalvatzaki, Jan Peters, Kristian Kersting

This preprint introduces NEUPRO, a neuro-symbolic safety system that learns reusable visual safety predicates from human-written symbolic rules through a differentiable graph reasoner. On the REASON real-robot benchmark it reaches 92% mean safety-classification accuracy across seven tasks versus 55% for Qwen3.5 and 44% for DeepSeek-VL2, and it can explain violations in first-order logic. The result matters because it offers a path to interpretable, transferable safety constraints instead of opaque per-task cost functions.

Read the full paper (PDF)
arXiv
SafetyExpert

Belief-Aware Multi-Agent Path Finding under Map Uncertainty

Viraj Parimi, Shao-Hung Chan, Han Zhang, Jingkai Chen, Brian Williams

Multi-Agent Path Finding (MAPF) aims to find collision-free paths for multiple agents in a shared environment. Classical MAPF assumes that all static obstacles are known in advance, but real-world environments can change unexpectedly due to fallen objects, spills, or other local disturbances.

Read the full paper (PDF)
arXiv
SafetyExpert

ExceptionDrive: A Planning-Oriented Counterfactual Corner-Case Benchmark for Autonomous Driving

Ziyi Luo, Zhe Sun, Yehao Lu, Lei Zhou, Lisheng Wu, Xuewei Li, Zequn Qin, Xi Li

Average performance on routine driving benchmarks does not establish planner reliability under rare, safety-critical hazards. We proposed ExceptionDrive, a counterfactual planning benchmark that uses VLM-assisted screening, localized multi-view editing, and quality auditing to insert hazards into real nuScenes scenes while preserving their context.

Read the full paper (PDF)

Topic pages

Manipulation · 20VLA · 16Navigation · 15Perception · 9General robotics · 9Robot learning · 6Humanoid · 5Safety · 3Simulation · 3

Papers come from arXiv robotics feeds. Where our summary is not written yet, you see the opening lines of the abstract.