DrPlanner: Diagnosis and Repair of Motion Planners for Automated Vehicles Using Large Language Models

被引:0
作者
Lin, Yuanfei [1 ]
Li, Chenran [2 ]
Ding, Mingyu [2 ]
Tomizuka, Masayoshi [2 ]
Zhan, Wei [2 ]
Althoff, Matthias [1 ]
机构
[1] Tech Univ Munich, Sch Computat Informat & Technol, D-85748 Garching, Germany
[2] Univ Calif Berkeley, Dept Mech Engn, Berkeley, CA 94720 USA
来源
IEEE ROBOTICS AND AUTOMATION LETTERS | 2024年 / 9卷 / 10期
关键词
Automated software repair; integrated planning and learning; intelligent transportation systems; large language models; motion and path planning;
D O I
10.1109/LRA.2024.3441493
中图分类号
TP24 [机器人技术];
学科分类号
080202 ; 1405 ;
摘要
Motion planners are essential for the safe operation of automated vehicles across various scenarios. However, no motion planning algorithm has achieved perfection in the literature, and improving its performance is often time-consuming and labor-intensive. To tackle the aforementioned issues, we present ${\mathtt {DrPlanner}}$, the first framework designed to automatically diagnose and repair motion planners using large language models. Initially, we generate a structured description of the planner and its planned trajectories from both natural and programming languages. Leveraging the profound capabilities of large language models, our framework returns repaired planners with detailed diagnostic descriptions. Furthermore, our framework advances iteratively with continuous feedback from the evaluation of the repaired outcomes. Our approach is validated using both search- and sampling-based motion planners for automated vehicles; experimental results highlight the need for demonstrations in the prompt and show the ability of our framework to effectively identify and rectify elusive issues.
引用
收藏
页码:8218 / 8225
页数:8
相关论文
共 54 条
  • [11] Feng ZY, 2020, FINDINGS OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS, EMNLP 2020, P1536
  • [12] Fried D., 2023, P INT C LEARN REPR
  • [13] Jain N., 2024, P INT C LEARN REPR
  • [14] CURE: Code-Aware Neural Machine Translation for Automatic Program Repair
    Jiang, Nan
    Lutellier, Thibaud
    Tan, Lin
    [J]. 2021 IEEE/ACM 43RD INTERNATIONAL CONFERENCE ON SOFTWARE ENGINEERING (ICSE 2021), 2021, : 1161 - 1173
  • [15] Kiciman Emre, 2023, Causal reasoning and large language models: Opening a new frontier for causality
  • [16] Scenario Factory: Creating Safety-Critical Traffic Scenarios for Automated Vehicles
    Klischat, Moritz
    Liu, Edmond Irani
    Hoeltke, Fabian
    Althoff, Matthias
    [J]. 2020 IEEE 23RD INTERNATIONAL CONFERENCE ON INTELLIGENT TRANSPORTATION SYSTEMS (ITSC), 2020,
  • [17] Kojima T, 2022, ADV NEUR IN
  • [18] Kolak R., 2022, INT C LEARN REPR DEE
  • [19] FixMiner: Mining relevant fix patterns for automated program repair
    Koyuncu, Anil
    Liu, Kui
    Bissyande, Tegawende F.
    Kim, Dongsun
    Klein, Jacques
    Monperrus, Martin
    Le Traon, Yves
    [J]. EMPIRICAL SOFTWARE ENGINEERING, 2020, 25 (03) : 1980 - 2024
  • [20] Code as Policies: LanguageModel Programs for Embodied Control
    Liang, Jacky
    Huang, Wenlong
    Xia, Fei
    Xu, Peng
    Hausman, Karol
    Ichter, Brian
    Florence, Pete
    Zeng, Andy
    [J]. 2023 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION (ICRA 2023), 2023, : 9493 - 9500