Large Language Model-Assisted Preparation of Transportation Management Plans: A Case Study with WisDOT WisTMP System
Summary
本文提出一个基于大语言模型的框架，利用微调的开源 LLM 自动化生成交通管理计划（TMP），并在威斯康星州交通部的 WisTMP 系统中进行案例研究，发现微调显著提升性能，但模型仍难以生成项目特定理由和准确的成本估算。
View Cached Full Text
Cached at: 10/09/26, 06:48 AM
# Large Language Model-Assisted Preparation of Transportation Management Plans: A Case Study with WisDOT WisTMP System Source: [https://arxiv.org/abs/2610.10650](https://arxiv.org/abs/2610.10650) [View PDF](https://arxiv.org/pdf/2610.10650) > Abstract:Work zones are critical yet hazardous components of transportation infrastructure, requiring carefully designed Transportation Management Plans \(TMPs\) to ensure safety and mobility\. However, TMP preparation remains labor\-intensive and heavily dependent on practitioner expertise\. This paper proposes a Large Language Model \(LLM\)\-assisted framework to automate TMP content generation, leveraging the WisDOT WisTMP system as the application context\. The framework fine\-tunes multiple open\-source LLMs across different model scales and deploys them locally to ensure data security\. To support model training, we construct a domain\-specific dataset from historical WisTMP documents by converting PDF files into structured question\-answer pairs in JSON format\. Experimental results show that fine\-tuning significantly improves performance across standard text generation metrics\. Further section\-wise and strategy\-level analyses reveal that, while LLMs achieve strong overall performance, they tend to over\-generate strategies and struggle to produce project\-specific justifications and accurate cost estimates\. In addition, scaling from 7B/8B to 14B yields limited gains\. These findings demonstrate the potential of LLMs to improve TMP preparation efficiency while highlighting remaining challenges in LLM\-assisted TMP development\. The source code and demo videos will be publicly available at[this https URL](https://zihaosheng.github.io/TMP-LLM/)\. ## Submission history From: Zihao Sheng \[[view email](https://arxiv.org/show-email/fca47c40/2610.10650)\] **\[v1\]**Wed, 7 Oct 2026 15:51:01 UTC \(1,682 KB\)
Similar Articles
Cost-Optimal Foundation Model Deployment Portfolio for Transportation Management
This paper formulates the Foundation Model Deployment Portfolio problem to minimize total cost of ownership for deploying LLMs and VLMs across transportation management functions, considering quality, latency, and safety constraints. A case study shows cost savings of 97% compared to all-closed-API baselines.
TransitLM: A Large-Scale Dataset and Benchmark for Map-Free Transit Route Generation
This paper introduces TransitLM, a large-scale dataset of over 13 million transit route planning records from Chinese cities, enabling map-free route generation via LLMs trained directly on planning data.
MetroLLM-Bench: Evaluating Language Models as Transit Kiosk Runtimes
MetroLLM-Bench is a 955-case benchmark for evaluating language models as transit kiosk policy layers, showing that small fine-tuned models can match larger models on structured tool-use and fare-quoting tasks.
Plan-and-Patch: Diffusion Language Models for Agentic Planning
提出 Plan-and-Patch 框架，利用扩散语言模型生成结构化计划并通过局部修补（保留前后缀、重新生成受影响区域）实现高效计划修复；在 Natural Plan 上无任务训练时修复成功率约为自回归模型的两倍，并将计划生成延迟降低 39-46%。
ChatPlanner: A Large Language Model Framework for Personalized Public Transit Routing
ChatPlanner is a novel framework that uses fine-tuned LLMs with Retrieval-Augmented Generation (RAG) to interpret user preferences from natural language queries and integrate them into public transit routing algorithms, outperforming existing route planners.