BioVLN is a simulation platform for visual-language navigation in biomedical laboratories, where robots must approach instruments from an usable operating side while maintaining clearance from nearby equipment. It represents each instrument through three regions: the physical body, a clearance region, and an operation area. The representation is used consistently for scene generation, target placement, navigation evaluation, and safety analysis. The platform includes 47 scenes and 1,667 episodes. According to the abstract, geometric exploration achieves 74.4–87.5% success, while sampling multiple valid positions in the operation area raises success to 83.3–92.5% and reduces unsafe proximity.
No heat snapshots are available in the last 24 hours.