A Review of Learning-Based Motion Planning: Toward a Data-Driven Optimal Control Approach

2025-12-12

AI Summary

Advanced autonomous driving motion planning faces a fundamental trade-off between interpretability and adaptability: conventional optimal control methods are transparent yet lack generalizability, whereas end-to-end learning approaches are robust but unverifiable. To resolve this, we propose a data-driven optimal control paradigm that synergizes the verifiability of control theory with the adaptability of machine learning, continuously refining dynamics models, cost functions, and safety constraints using real-world data. Our framework introduces three novel capabilities—human-centered customization, platform-adaptive dynamics, and system self-optimization—establishing the first unified, interpretable, verifiable, and evolvable learning-based motion planning architecture. It integrates optimal control, imitation learning, reinforcement learning, generative AI, and online system identification, supporting both end-to-end and modular co-modeling. Experiments demonstrate significant improvements in planning safety and cross-scenario generalization, enabling human–machine collaborative, interpretable decision-making—providing theoretical foundations and technical pathways for trustworthy intelligent transportation systems.

Technology Category

• Humans and AI: Human-Aware Planning and Behavior Prediction
• Planning, Routing, and Scheduling: Planning/Scheduling and Learning
• Intelligent Robots: Motion and Path Planning

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Motion planning for high-level autonomous driving is constrained by a fundamental trade-off between the transparent, yet brittle, nature of pipeline methods and the adaptive, yet opaque, "black-box" characteristics of modern learning-based systems. This paper critically synthesizes the evolution of the field -- from pipeline methods through imitation learning, reinforcement learning, and generative AI -- to demonstrate how this persistent dilemma has hindered the development of truly trustworthy systems. To resolve this impasse, we conduct a comprehensive review of learning-based motion planning methods. Based on this review, we outline a data-driven optimal control paradigm as a unifying framework that synergistically integrates the verifiable structure of classical control with the adaptive capacity of machine learning, leveraging real-world data to continuously refine key components such as system dynamics, cost functions, and safety constraints. We explore this framework's potential to enable three critical next-generation capabilities: "Human-Centric" customization, "Platform-Adaptive" dynamics adaptation, and "System Self-Optimization" via self-tuning. We conclude by proposing future research directions based on this paradigm, aimed at developing intelligent transportation systems that are simultaneously safe, interpretable, and capable of human-like autonomy.

Problem

• Resolving the trade-off between transparent but brittle pipeline methods and adaptive but opaque learning-based systems in autonomous driving motion planning.
• Proposing a data-driven optimal control framework that integrates classical control's verifiability with machine learning's adaptability.
• Enabling next-generation capabilities like human-centric customization, platform-adaptive dynamics, and system self-optimization for trustworthy autonomous systems.

Innovation

• Data-driven optimal control integrates classical control with machine learning
• Leverages real-world data to refine system dynamics and safety constraints
• Enables human-centric customization and platform-adaptive self-optimization