Agentic Leave and Dispatch Automation for Trucking Fleets Using MCP and LLMs
DOI:
https://doi.org/10.63282/3050-9416.IJAIBDCMS-V7I1P111Keywords:
Large Language Models (Llms), Python, Flutter, Model Context Protocol, Postgres, REST API, Conversational Bot, Lang Graph, Tool CallingAbstract
Large trucking companies employing thousands of drivers operate in a complex work pattern with weekend and night shift operations. These kinds of large fleets face operational disruptions due to unplanned absences by the drivers. Looking up the leave balances in the HR system manually and rescheduling dispatches are slow, error-prone, and often result in unexpected payroll deductions due to insufficient leave balances. This paper presents Conversational Dispatcher Bot installed as an app on mobile phone which leverages Large Language Models (LLM) to understand driver queries and use Model Context Protocol(MCP) for secure enterprise tool calling. Using MCP server bot connects to HR system to check leave balances, leave eligibility, estimates the payroll deductions, process the leave and connects to Scheduling system to reassign affected pickups/drop-offs to the alternate qualified drivers. The proposed architecture transforms driver’s phone into operational control interface reducing the dispatcher workload immensely, improving scheduling continuity, increasing policy transparency and lowering driver attrition through fair and explainable decision-making.
References
1. Procter, D., & Sousa Jr, P. (2021). Goldilocks and the Three Dispatchers: Quantifying the Impact of Dispatcher Management on Truck Driver Performance.
2. Agarwal, K., Ananthanarayanan, S., & Srinivasan, S. (2024). Enhancing iot based plant health monitoring through advanced human plant interaction using large language models and mobile applications. arXiv preprint arXiv:2409.15910.
3. Sapkota, R., Shrestha, R., Rijal, M., & Karkee, M. LangChain vs. LangGraph vs. LangSmith: Taxonomies of Agentic AI Toolchains for End-to-End Orchestration. Authorea Preprints.
4. Jeong, C. (2025). A Study on the MCP x A2A Framework for Enhancing Interoperability of LLM-based Autonomous Agents. arXiv preprint arXiv:2506.01804.
5. Agents - Docs by LangChain
6. Ding, P. (2025). Toolregistry: A protocol-agnostic tool management library for function-calling llms. arXiv preprint arXiv:2507.10593.
7. Model Context Protocol (MCP) - Docs by LangChain.
