SensorMCP: A Model Context Protocol Server for Custom Sensor Tool Creation

The rising demand for customized sensor systems, such as wildlife and urban monitoring, underscores the need for scalable, AI-driven solutions. The Model Context Protocol (MCP) enables large language models (LLMs) to interface with external tools, yet lacks automated sensor tool generation. We propose SensorMCP, a novel MCP server framework that enables LLMs to dynamically generate and operate sensor tools through a tool-language co-development pipeline. Our contributions include: (1) a SensorMCP architecture for automated tool and language co-evolution, (2) an automated sensor toolbox generating tailored tools, and (3) language assets producing tool descriptions and linguistic modules. A preliminary evaluation using real-world zoo datasets demonstrates the practicality and efficiency of SensorMCP, achieving up to 95% tool success rate in scenarios like animal monitoring. This work advances sensor systems by pioneering the co-evolution of LLMs and sensor tools, offering a scalable framework for customized sensing in mobile systems. The source code and dataset are publicly available at https://sensormcp.github.io/sensor-mcp/.