Fujitsu Sdt MCP

🚀 Fujitsu Social Digital Twin MCP Server

This project aims to seamlessly integrate Fujitsu's social digital twin technology with the Model Context Protocol (MCP) to enhance the capabilities of Large Language Models (LLMs). The system supports various operations, including simulated data management, scenario analysis, and natural language generation.

🚀 Quick Start

This server enhances LLM functionality by integrating social digital twin technology with MCP. It supports diverse operations and offers features like multi - dimensional simulation and intelligent data processing.

✨ Features

• Multi - dimensional Simulation: Covers multiple domains such as traffic flow, crowd movement, and social activities.
• Intelligent Data Processing: Can automatically parse and process large - scale structured and unstructured data.
• Dynamic Interaction Capability: Supports real - time queries, predictive analysis, and adaptive adjustment functions.

📦 Installation

Prerequisites

1. Operating System: Linux 64 - bit or Windows 10/11 (Home Edition and above only).
2. Memory: At least 8GB, 16GB is recommended.
3. Disk Space: At least 50GB of available space.
4. Installation Dependencies:
   • Python 3.8 or higher.
   • pip package manager.
   • It is recommended to install GPU acceleration libraries (e.g., CUDA).

Environment Preparation

1. Update system software and install necessary development tools.
2. Configure environment variables:

   export FUJITSU_SDT_HOME=/path/to/sdt - installation
   

3. Add the Fujitsu Social Digital Twin SDK to the PATH environment variable.

Installation Steps

1. Download and extract the installation package.
2. Install dependencies using pip:

   python -m pip install -r requirements.txt
   

3. Initialize the database and configuration files.

💻 Usage Examples

Basic Usage

Resource Access

• resource://simulation_overview: Overview of digital twin technology
• resource://metrics_explanation: Explanation of simulation metrics
• resource://scenario_samples: Scenario examples

Tool Usage

• list_simulations: Get the list of simulations
• start_simulation: Start a simulation
• get_result: Retrieve simulation results
• fetch_metrics: Get simulation metrics
• list_data: Retrieve the list of simulation data
• retrieve_data: Get specific simulation data
• analyze_traffic: Analyze traffic flow
• compare_scenarios: Compare different scenarios
• generate_config: Generate configuration based on natural language

Advanced Usage

Get the list of simulations

tool list_simulations

Start a simulation

tool start_simulation {"simdataId": "your_id"}

Analyze traffic flow

tool analyze_traffic {"simulationId": "id", "区域": "东京", "时间范围": "早高峰", "情景": "交通优化"}

Generate configuration based on natural language

tool generate_config {"描述": "我要在东京的早高峰时段运行交通模拟"}

Example Dialogue

User: I want to run a traffic simulation during the morning rush hour in Tokyo.

System: I can help you set up a traffic simulation during the morning rush hour in Tokyo. Let's first generate a configuration based on your natural language description:

[Tool Execution: generate_config]

Generated Configuration:

{
  "类型": "交通",
  "参数": {
    "区域": "东京",
    "时间范围": "早高峰"
  },
  "名称": "traffic_20250416_123045"
}

Next, I will list the available simulation data:

[Tool Execution: list_data]

From this data, I selected the data ID suitable for the morning rush - hour traffic conditions in Tokyo and started the simulation:

[Tool Execution: start_simulation]

The simulation has been started with the ID "sim_12345". It will take some time to complete. I will analyze the results after they are generated.

(After the simulation is completed) Let's analyze the results:

[Tool Execution: analyze_traffic]

Analysis Results:

• CO2 emissions: 1,250kg
• Average travel time: 42 minutes
• Total traffic flow: 15,700 vehicles

The results show that the congestion on the route from Shibuya to Shinjuku is particularly severe. Here are some improvement suggestions:

1. Optimize traffic signal control.
2. Implement one - way traffic policies for specific roads and time periods.
3. Increase the frequency of public transportation.

Do you need to view a more detailed analysis?

📚 Documentation

Contribution

Please report issues or suggest features through GitHub Issues. Pull Requests are welcome.

License

This project is released under the MIT License. Please refer to the LICENSE file for details.

Acknowledgments

Thank Fujitsu for providing technical support, and special thanks to the development team for their contributions.