Self-Supervised Path Planning in Unstructured Environments via Global-Guided Differentiable Hard Constraint Projection

2026-01-27

AI Summary

This work proposes a self-supervised path planning framework to address the challenges of ensuring safety, coping with scarce labeled data, and operating under limited computational resources in unstructured environments—conditions under which existing methods struggle to balance real-time performance and path feasibility. The approach innovatively integrates a globally guided artificial potential field (G-APF) with a differentiable hard-constraint projection layer, embedding actuator limits and geometric constraints directly into the network output as hard constraints without requiring human annotations. This guarantees that generated trajectories remain on the feasible manifold and satisfy dynamic constraints. Evaluated across 20,000 test scenarios, the method achieves an 88.75% success rate; CARLA closed-loop simulations confirm physical realizability, and inference latency of only 94 ms on an NVIDIA Jetson Orin NX demonstrates its suitability for real-time embedded deployment.

Technology Category

• Planning, Routing, and Scheduling: Planning with Language Models
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Intelligent Robots: Motion and Path Planning

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web
• Search and Retrieval-Augmented AI: Agentic search

Abstract

Deploying deep learning agents for autonomous navigation in unstructured environments faces critical challenges regarding safety, data scarcity, and limited computational resources. Traditional solvers often suffer from high latency, while emerging learning-based approaches struggle to ensure deterministic feasibility. To bridge the gap from embodied to embedded intelligence, we propose a self-supervised framework incorporating a differentiable hard constraint projection layer for runtime assurance. To mitigate data scarcity, we construct a Global-Guided Artificial Potential Field (G-APF), which provides dense supervision signals without manual labeling. To enforce actuator limitations and geometric constraints efficiently, we employ an adaptive neural projection layer, which iteratively rectifies the coarse network output onto the feasible manifold. Extensive benchmarks on a test set of 20,000 scenarios demonstrate an 88.75\% success rate, substantiating the enhanced operational safety. Closed-loop experiments in CARLA further validate the physical realizability of the planned paths under dynamic constraints. Furthermore, deployment verification on an NVIDIA Jetson Orin NX confirms an inference latency of 94 ms, showing real-time feasibility on resource-constrained embedded hardware. This framework offers a generalized paradigm for embedding physical laws into neural architectures, providing a viable direction for solving constrained optimization in mechatronics. Source code is available at: https://github.com/wzq-13/SSHC.git.

Problem

• autonomous navigation
• unstructured environments
• safety
• data scarcity
• computational constraints

Innovation

• Self-supervised learning
• Differentiable hard constraint projection
• Global-Guided Artificial Potential Field
• Embedded autonomous navigation
• Feasible manifold projection