WildOS: Open-Vocabulary Object Search in the Wild(NPO-53765-1)

aeronautics
WildOS: Open-Vocabulary Object Search in the Wild
(NPO-53765-1)
Overview
This work presents WildOS, a unified system for long-range, open-vocabulary object search that combines safe geometric exploration with semantic visual reasoning. WildOS builds a sparse navigation graph to maintain spatial memory, while utilizing a foundation-model-based vision module, ExploRFM, to score frontier nodes. ExploRFM simultaneously predicts traversability, visual frontiers, and object similarity in image space, enabling real-time, onboard semantic navigation tasks. The resulting vision-scored graph enables the robot to explore semantically meaningful directions while ensuring geometric safety. Furthermore, we introduce a particle-filter-based localization method that estimates coarse target positions beyond the robot's immediate depth horizon, enabling effective planning toward distant goals. Extensive closed-loop field experiments across diverse off-road and urban terrains demonstrate that WildOS enables robust generalization, significantly outperforming purely geometric and purely vision-based baselines in both efficiency and autonomy. Our results highlight the potential of vision foundation models to drive open-world robotic behaviors that are both semantically informed and geometrically grounded.
Software Details

Category
Aeronautics
Reference Number
NPO-53765-1
Release Type
Open Source
Operating System
Contact Us About This Technology

Stay up to date, follow NASA's Technology Transfer Program on:
facebook twitter linkedin youtube
Facebook Logo X Logo Linkedin Logo Youtube Logo