Schematic maps improve robot navigation to objects indoors
ScheNav: Schematic-Guided Object Goal Navigation
Robotics
Summary
Finding objects quickly inside complex buildings is hard for robots, but humans use simple maps and sketches to help. The authors created a method where robots learn to use rough floorplans as guidance to explore and find target objects better. They made a big dataset from floorplans to train the robots and a system that combines the sketch guidance with what the robot sees as it moves. Tests showed robots navigated more successfully and efficiently using these rough maps, even when the maps were incomplete or noisy.
What this means in practice
- •For indoor robotics developers: Improve robot navigation inside buildings by integrating rough schematic maps with visual exploration for more efficient object finding.
- •For augmented reality app builders: Create AR tools that overlay simplified floorplan directions to guide users in unfamiliar indoor spaces.
Authors
Guichen Huang, Chenchen Jing, Che Sun, Yuwei Wu, Yunde Jia
Abstract
Object goal navigation requires an embodied agent to locate and reach target objects in complex environments. Humans often rely on low-cost yet informative schematic maps, such as rough sketches or directional annotations. We develop a workflow for schematic-guided embodied navigation with two components: a training-data generation pipeline and ScheNav, a closed-loop navigation framework. Our lightweight data-generation pipeline turns 2D floorplans into navigation training environments, yielding Furnished-RPLAN, a dataset of over 33k furnished layouts with paired navigation episodes and noise-controllable schematic maps. ScheNav turns rough schematic guidance into exploration and stopping decisions by using SchemaField, a learned spatial module that fuses the guidance with accumulated observations and predicts spatial value fields. The framework updates these fields as exploration proceeds, with trajectory-based training and a cross-frame consistency objective supporting coherent guidance across changing viewpoints. ScheNav learns to use schematic guidance entirely from simulated floorplans and transfers this ability to unseen embodied environments. Experiments show improved navigation success and path efficiency across schematic maps, goal ellipses, and pointed directions, together with tolerance to the tested schematic perturbations and omissions. Together, these results demonstrate that rough schematic guidance offers a practical way to improve embodied navigation.