Acting from Belief, Looking When Needed: A Bayesian Spatial World Model for Navigation under Intermittent Perception

2026-10-08

AI Summary

This study addresses navigation observation disruptions caused by sensor sharing and the reliance on high-frequency perception by proposing ALONE, a Bayesian spatial world model. The method innovatively decouples perception from decision-making by integrating geometric priors, reliability prediction maps, and motion planning modules. This architecture enables agents to act upon internal beliefs and request observations only when reliability is insufficient, thereby achieving robust navigation under intermittent perception. In closed-loop simulations, the approach attains success rates of 97%–98% while requiring an observation rate of merely approximately 1%. Furthermore, real-world indoor flight experiments achieve complete success across all trials, validating its efficient and low-overhead navigation capabilities.

Abstract

Robot navigation commonly uses wide-coverage, high-frequency sensing to reduce partial observability; this reliance becomes restrictive when another task temporarily redirects a shared sensor from navigation, interrupting navigation-relevant observations. We study navigation under intermittent perception: acting from an internal spatial belief and looking again only when execution needs a new observation, potentially freeing the shared sensor for other tasks between navigation observations. ALONE, a Bayesian spatial world model, propagates a structured spatial belief using executed actions and corrects it with selectively acquired observations; learned priors over common geometric structures infer unobserved structure from available observation history. It decodes the belief into a spatial estimate for the motion-planning module and predicts a reliability map expressing confidence in the estimate's accuracy. ALONE requests an observation only if insufficient reliability hinders navigation and new evidence should make relevant-region spatial information more reliable; otherwise, it continues acting from the propagated belief. We instantiate ALONE for drone navigation with intermittent single-camera depth images. Across two simulated scene families, it achieves 98% and 97% closed-loop success at a 10 Hz decision rate. Among successful trials, median fractions of decision steps requiring a new depth observation are only 0.9% and 1.3%, respectively, demonstrating high navigation success with substantially reduced observation demand. Real-world indoor flight experiments further validate navigation under intermittent depth observations, with all 10 trials successful.

Problem

• robot navigation
• intermittent perception
• partial observability
• shared sensor
• spatial belief

Innovation

• Bayesian spatial world model
• intermittent perception
• structured spatial belief
• reliability map
• selective observation