Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Autonomous Driving in Unstructured Environments: How Far Have We Come?

Chen Min, Shubin Si, Xu Wang, Hanzhang Xue, Weizhong Jiang, Zitong Chen, Mengmeng Li, Jilin Mei, Erke Shang, Zhipeng Xiao, Bin Dai, Qi Zhu, Hao Fu, Dawei Zhao, Liang Xiao, Yiming Nie, Yu Hu

TL;DR

The paper surveys autonomous driving in unstructured outdoor environments, addressing the unique challenges of off-road, rural, and rugged terrains. It reviews more than 250 papers across offline mapping, pose estimation, environmental perception, path planning, motion control, end-to-end driving, and datasets, emphasizing LiDAR-based and fusion-based approaches. It discusses current limitations, trends, and future directions, including high-precision dynamic mapping, multimodal pose estimation, robust perception, adaptive planning, and edge computing. The work also provides an active literature repository to support ongoing research.

Abstract

Research on autonomous driving in unstructured outdoor environments is less advanced than in structured urban settings due to challenges like environmental diversities and scene complexity. These environments-such as rural areas and rugged terrains-pose unique obstacles that are not common in structured urban areas. Despite these difficulties, autonomous driving in unstructured outdoor environments is crucial for applications in agriculture, mining, and military operations. Our survey reviews over 250 papers for autonomous driving in unstructured outdoor environments, covering offline mapping, pose estimation, environmental perception, path planning, end-to-end autonomous driving, datasets, and relevant challenges. We also discuss emerging trends and future research directions. This review aims to consolidate knowledge and encourage further research for autonomous driving in unstructured environments. To support ongoing work, we maintain an active repository with up-to-date literature and open-source projects at: https://github.com/chaytonmin/Survey-Autonomous-Driving-in-Unstructured-Environments.

Autonomous Driving in Unstructured Environments: How Far Have We Come?

TL;DR

The paper surveys autonomous driving in unstructured outdoor environments, addressing the unique challenges of off-road, rural, and rugged terrains. It reviews more than 250 papers across offline mapping, pose estimation, environmental perception, path planning, motion control, end-to-end driving, and datasets, emphasizing LiDAR-based and fusion-based approaches. It discusses current limitations, trends, and future directions, including high-precision dynamic mapping, multimodal pose estimation, robust perception, adaptive planning, and edge computing. The work also provides an active literature repository to support ongoing research.

Abstract

Research on autonomous driving in unstructured outdoor environments is less advanced than in structured urban settings due to challenges like environmental diversities and scene complexity. These environments-such as rural areas and rugged terrains-pose unique obstacles that are not common in structured urban areas. Despite these difficulties, autonomous driving in unstructured outdoor environments is crucial for applications in agriculture, mining, and military operations. Our survey reviews over 250 papers for autonomous driving in unstructured outdoor environments, covering offline mapping, pose estimation, environmental perception, path planning, end-to-end autonomous driving, datasets, and relevant challenges. We also discuss emerging trends and future research directions. This review aims to consolidate knowledge and encourage further research for autonomous driving in unstructured environments. To support ongoing work, we maintain an active repository with up-to-date literature and open-source projects at: https://github.com/chaytonmin/Survey-Autonomous-Driving-in-Unstructured-Environments.

Paper Structure

This paper contains 63 sections, 3 equations, 10 figures, 18 tables.

Table of Contents

  1. Offline Mapping
  2. Challenges of Offline Mapping in Unstructured Environments
  3. LiDAR-based Methods
  4. Fusion-based Methods
  5. Pose Estimation
  6. Challenges of Pose Estimation in Unstructured Environments
  7. Matching-based Methods
  8. Matching with High-Definition Map
  9. Matching with Lightweight Map
  10. Place Recognition / Relocalization
  11. Odometry-based Methods
  12. Environmental Perception
  13. Challenges of Environmental Perception in Unstructured Environment
  14. LiDAR-based Methods
  15. Traversability Estimation
  16. ...and 48 more sections

Figures (10)

  • Figure 1: Inherent characteristics of unstructured environments. Compared to structured environments, autonomous driving in unstructured environment encounters various challenges, including the diversity of environmental types, a vast array of elements, ambiguity in semantic categories, disorderly scene structures, complex road conditions, and vulnerability of satellite signals.
  • Figure 2: Taxonomy of autonomous driving in unstructured environments.
  • Figure 3: Typical components of the autonomous driving system. The modules for autonomous driving systems in both unstructured and structured environments are fundamentally similar, comprising offline mapping, pose estimation, environmental perception, path planning, and motion control.
  • Figure 4: Challenges faced by autonomous driving systems in unstructured environments.
  • Figure 5: SLAM-based pose estimation methods towards unstructured environments.
  • ...and 5 more figures