Trajectory Prediction for Autonomous Driving: Progress, Limitations, and Future Directions

2025-03-05

AI Summary

This paper addresses the core challenges of inaccurate and unsafe trajectory prediction for surrounding agents in dynamic autonomous driving environments. It systematically surveys mainstream approaches from 2014 to 2024 and proposes the first comprehensive taxonomy framework covering the entire technical stack—unifying definitions of input/output modalities, feature representation paradigms, and prediction architectures. Through comparative analysis of physics-based models, machine learning (ML), graph neural networks (GNNs), generative modeling (GANs and diffusion models), and multimodal fusion techniques, the study identifies six critical research gaps, including weak multi-agent interaction modeling and insufficient long-horizon generalization. The work clarifies methodological evolution trajectories and empirical performance boundaries, providing principled guidance for algorithm selection. Ultimately, it advances the development of next-generation trajectory prediction systems that are more robust, interpretable, and semantically grounded in traffic rules and human driving behavior.

Technology Category

• Humans and AI: Human-Aware Planning and Behavior Prediction
• Planning, Routing, and Scheduling: Model-Based Reasoning
• Multiagent Systems: Adversarial Agents

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Social Networks and Social Media: Influence propagation, information diffusion, and the prediction on networks

Abstract

As the potential for autonomous vehicles to be integrated on a large scale into modern traffic systems continues to grow, ensuring safe navigation in dynamic environments is crucial for smooth integration. To guarantee safety and prevent collisions, autonomous vehicles must be capable of accurately predicting the trajectories of surrounding traffic agents. Over the past decade, significant efforts from both academia and industry have been dedicated to designing solutions for precise trajectory forecasting. These efforts have produced a diverse range of approaches, raising questions about the differences between these methods and whether trajectory prediction challenges have been fully addressed. This paper reviews a substantial portion of recent trajectory prediction methods and devises a taxonomy to classify existing solutions. A general overview of the prediction pipeline is also provided, covering input and output modalities, modeling features, and prediction paradigms discussed in the literature. In addition, the paper discusses active research areas within trajectory prediction, addresses the posed research questions, and highlights the remaining research gaps and challenges.

Problem

• Ensuring safe navigation for autonomous vehicles in dynamic environments.
• Accurately predicting trajectories of surrounding traffic agents to prevent collisions.
• Reviewing and classifying recent trajectory prediction methods to address research gaps.

Innovation

• Taxonomy for classifying trajectory prediction methods
• Review of input/output modalities and modeling features
• Identification of research gaps in trajectory prediction