# Work Zones challenge VLM Trajectory Planning: Toward Mitigation and Robust Autonomous Driving

arXiv: https://arxiv.org/abs/2510.02803 · HTML: https://mlsanity.com/paper/2510.02803/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Placed in Chain-of-thought reasoning for driving, one of 17 directions in LLM reasoning.
- Builds on 18 of its 36 references that are in the map.
- First posted on arXiv 3 Oct 2025.

## Placed in

- LLM reasoning / [Chain-of-thought reasoning for driving](https://mlsanity.com/lr/chain-of-thought-reasoning-for-driving/)
- Multimodal LLMs / vision-language models / [Driving VLMs and open-vocabulary detection](https://mlsanity.com/mllm/driving-vlms-and-open-vocabulary-detection/)
- End-to-end autonomous driving / [VLMs for autonomous driving](https://mlsanity.com/driving/vlms-for-autonomous-driving/)

## Builds on (most cited first, 18 of 18)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Qwen2.5-VL Technical Report](https://mlsanity.com/paper/2502.13923/) (2025)
- [nuScenes: A multimodal dataset for autonomous driving](https://mlsanity.com/paper/1903.11027/) (2019)
- [Towards Robust Monocular Depth Estimation: Mixing Datasets for Zero-shot Cross-dataset Transfer](https://mlsanity.com/paper/1907.01341/) (2019)
- [VAD: Vectorized Scene Representation for Efficient Autonomous Driving](https://mlsanity.com/paper/2303.12077/) (2023)
- [DriveLM: Driving with Graph Visual Question Answering](https://mlsanity.com/paper/2312.14150/) (2023)
- [GPT-Driver: Learning to Drive with GPT](https://mlsanity.com/paper/2310.01415/) (2023)
- [Scaling Up Vision-Language Pre-training for Image Captioning](https://mlsanity.com/paper/2111.12233/) (2021)
- [Senna: Bridging Large Vision-Language Models and End-to-End Autonomous Driving](https://mlsanity.com/paper/2410.22313/) (2024)
- [Vision Language Models in Autonomous Driving: A Survey and Outlook](https://mlsanity.com/paper/2310.14414/) (2023)
- [Dolphins: Multimodal Language Model for Driving](https://mlsanity.com/paper/2312.00438/) (2023)
- [CODA: A Real-World Road Corner Case Dataset for Object Detection in Autonomous Driving](https://mlsanity.com/paper/2203.07724/) (2022)
- [OpenEMMA: Open-Source Multimodal Model for End-to-End Autonomous Driving](https://mlsanity.com/paper/2412.15208/) (2024)
- [ROADWork: A Dataset and Benchmark for Learning to Recognize, Observe, Analyze and Drive Through Work Zones](https://mlsanity.com/paper/2406.07661/) (2024)
- [Q: How to Specialize Large Vision-Language Models to Data-Scarce VQA Tasks? A: Self-Train on Unlabeled Images!](https://mlsanity.com/paper/2306.03932/) (2023)
- [Can VLMs Detect and Localize Fine-Grained AI-Edited Images?](https://mlsanity.com/paper/2505.15644/) (2025)
- [Towards Stealthy and Effective Backdoor Attacks on Lane Detection: A Naturalistic Data Poisoning Approach](https://mlsanity.com/paper/2508.15778/) (2025)
- ["I Can See Forever!": Evaluating Real-time VideoLLMs for Assisting Individuals with Visual Impairments](https://mlsanity.com/paper/2505.04488/) (2025)

## Built on by (most cited first, 4 of 4)

- [UniDrive-WM: Unified Understanding, Planning and Generation World Model for Autonomous Driving](https://mlsanity.com/paper/2601.04453/) (2026)
- [GraphPilot: Grounded Scene Graph Conditioning for Language-Based Autonomous Driving](https://mlsanity.com/paper/2511.11266/) (2025)
- [TimeSpot: Benchmarking Geo-Temporal Understanding in Vision-Language Models in Real-World Settings](https://mlsanity.com/paper/2603.06687/) (2026)
- [WorkDrive: Roadwork Chain of Causation for Autonomous Driving](https://mlsanity.com/paper/2607.14727/) (2026)

## Abstract (arXiv)

Visual Language Models (VLMs), with powerful multimodal reasoning capabilities, are gradually integrated into autonomous driving by several automobile manufacturers to enhance planning capability in challenging environments. However, the trajectory planning capability of VLMs in work zones, which often include irregular layouts, temporary traffic control, and dynamically changing geometric structures, is still unexplored. To bridge this gap, we conduct the textitfirst systematic study of VLMs for work zone trajectory planning, revealing that mainstream VLMs fail to generate correct trajectories in 68.0% of cases. To better understand these failures, we first identify candidate patterns via subgraph mining and clustering analysis, and then confirm the validity of 8 common failure patterns through human verification. Building on these findings, we propose REACT-Drive, a trajectory planning framework that integrates VLMs with Retrieval-Augmented Generation (RAG). Specifically, REACT-Drive leverages VLMs to convert prior failure cases into constraint rules and executable trajectory planning code, while RAG retrieves similar patterns in new scenarios to guide trajectory generation. Experimental results on the ROADWork dataset show that REACT-Drive yields a reduction of around 3× in average displacement error relative to VLM baselines under evaluation with Qwen2.5-VL. In addition, REACT-Drive yields the lowest inference time (0.58s) compared with other methods such as fine-tuning (17.90s). We further conduct experiments using a real vehicle in 15 work zone scenarios in the physical world, demonstrating the strong practicality of REACT-Drive.