Implicit Dual-Control for Visibility-Aware Navigation in Unstructured Environments

2025-07-06

AI Summary

To address occlusions and unobserved regions arising from limited field-of-view in unknown unstructured environments for autonomous ground vehicles, this paper proposes a perception–control coupled two-layer planning framework. The method dynamically models perceptual uncertainty and implicitly embeds it into the Variational Autoencoder–Model Predictive Path Integral (VA-MPPI) optimization, enabling joint optimization of safety and navigation performance without explicit exploration objectives. By integrating a visibility-aware perception model with stochastic sampling–based model predictive control (MPC), the framework achieves online trade-offs between exploration and exploitation. Experimental evaluation across diverse off-road scenarios demonstrates an 84% task success rate and zero collision rate, significantly outperforming conventional deterministic controllers. The key contribution lies in the implicit incorporation of perceptual uncertainty into MPPI-based trajectory optimization, thereby unifying perception-aware decision-making and control under uncertainty.

Technology Category

• Planning, Routing, and Scheduling: Planning under Uncertainty
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Reasoning under Uncertainty: Sequential Decision Making

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Navigating complex, cluttered, and unstructured environments that are a priori unknown presents significant challenges for autonomous ground vehicles, particularly when operating with a limited field of view(FOV) resulting in frequent occlusion and unobserved space. This paper introduces a novel visibility-aware model predictive path integral framework(VA-MPPI). Formulated as a dual control problem where perceptual uncertainties and control decisions are intertwined, it reasons over perception uncertainty evolution within a unified planning and control pipeline. Unlike traditional methods that rely on explicit uncertainty objectives, the VA-MPPI controller implicitly balances exploration and exploitation, reducing uncertainty only when system performance would be increased. The VA-MPPI framework is evaluated in simulation against deterministic and prescient controllers across multiple scenarios, including a cluttered urban alleyway and an occluded off-road environment. The results demonstrate that VA-MPPI significantly improves safety by reducing collision with unseen obstacles while maintaining competitive performance. For example, in the off-road scenario with 400 control samples, the VA-MPPI controller achieved a success rate of 84%, compared to only 8% for the deterministic controller, with all VA-MPPI failures arising from unmet stopping criteria rather than collisions. Furthermore, the controller implicitly avoids unobserved space, improving safety without explicit directives. The proposed framework highlights the potential for robust, visibility-aware navigation in unstructured and occluded environments, paving the way for future advancements in autonomous ground vehicle systems.

Problem

• Navigating unknown unstructured environments with limited FOV
• Balancing exploration and exploitation in perception-control systems
• Reducing collision risks in occluded and cluttered scenarios

Innovation

• Visibility-aware model predictive path integral framework
• Implicitly balances exploration and exploitation
• Unified planning and control pipeline