LLM-Handover: Exploiting LLMs for Task-Oriented Handovers
Researchers introduced LLM-Handover, a framework that combines large language model reasoning with part segmentation to select context‑aware grasps for robot‑to‑human handovers. The system processes an RGB‑D image and a natural‑language task description, infers relevant object parts, and chooses grasps that maximize post‑handover usability. In zero‑shot hardware tests it achieved an 83% success rate across diverse tasks, and a user study showed participants preferred its handovers 86% of the time. The work appears in Robotics and Automation Letters.
High Precision Excavator Control
The video introduces an autonomous control stack designed for heavy‑duty excavator grading that delivers centimeter‑level surface accuracy, addressing the precision loss and torque under‑utilization of existing semi‑automatic systems. The solution splits into two modules: a hydraulics‑aware joint velocity controller that adapts...
Task-Oriented Robot-Human Handovers on Legged Manipulators
The researchers introduce AFT‑Handover, a framework that combines large language model‑driven affordance reasoning with texture‑based affordance transfer to enable zero‑shot, task‑oriented robot‑to‑human handovers. In a controlled user study, 71.43% of participants preferred AFT‑Handover over existing state‑of‑the‑art methods, citing reduced regrasping...
DiskChunGS: Large-Scale 3D Gaussian SLAM Through Chunk-Based Memory Management
DiskChunGS introduces a scalable 3D Gaussian splatting SLAM pipeline that overcomes traditional GPU memory constraints by treating scene reconstruction as a spatial streaming problem. The system partitions the environment into discrete chunks, keeping only the currently visible regions in GPU...