Skip to content

Latest commit

 

History

History
114 lines (72 loc) · 7.78 KB

README.md

File metadata and controls

114 lines (72 loc) · 7.78 KB

🤖 GridNLP

Introduction: - Substation Asset Maintenance Chatbot

GridNLP represents a conversational chat bot system engineered to provide indispensable support within the realm of Substation Asset Maintenance. By harnessing the capabilities of Natural Language Processing (NLP), GridNLP adeptly addresses user inquiries pertaining to maintenance procedures, assessment thresholds, issue troubleshooting, safety protocols, and equipment recommendations. Notably, GridNLP remains an ongoing development effort, dedicated to delivering precision, granularity, and insightful feedback to augment the maintenance protocols within the substation industry. The development framework for this endeavor is founded on tools such as NLTK, PyTorch, and FastAPI for the Python backend, complemented by the utilization of Flutter for mobile application development. The retrieval of responses is orchestrated through advanced techniques, including cosine similarity and tag matching. Additionally, a robust spell checker has been seamlessly integrated into the backend to ensure optimal response quality.

Image 1 Image 2

<<<<<<< HEAD GridNLP is an intelligent chatbot designed to assist users in the field of Substation Asset Maintenance. Leveraging natural language processing (NLP), it answers user queries related to maintenance tasks, including test procedures, acceptable limits, issue resolution, safety guidelines, and equipment recommendations. The chatbot aims to provide accurate, detailed, and informative responses to enhance the maintenance processes within the substation industry. The Project was build using Flutter for Mobile app development and FastAPI for python backend development. The models were created using pytorch and basic concepts of NLP. Intentation vs tag matching with cosine similarity have been added for better responses.

740981e0d3e30664abf875786a0f04a0c3c73476

Features

GridNLP offers a range of features to facilitate user interactions:

  1. 📚 Query Responses: The chatbot provides responses to user queries related to substation maintenance, covering various equipment classes such as transformers, circuit breakers, reactors, surge arrestors, and more.

  2. 📋 Procedure Guidance: GridNLP offers step-by-step guidance on maintenance procedures for different equipment classes.

  3. ⚠️ Safety Guidelines: Users can access safety guidelines to ensure secure maintenance practices, adhering to industry regulations and standards.

  4. 🔬 Test Procedures: The chatbot explains test procedures and acceptable limits for various equipment, allowing users to conduct tests effectively.

  5. 🛠️ Issue Resolution: GridNLP offers recommendations for resolving issues that may arise during maintenance tasks.

  6. 💡 Equipment Recommendations: Users can inquire about equipment recommendations based on their specific requirements.

  7. 🧠 Semantic Processing: The chatbot incorporates semantic processing to understand and respond to user queries more effectively.

  8. 🌐 Industry Standards: Users can access information on industry standards and regulations governing substation maintenance.

<<<<<<< HEAD To interact with GridNLP, you can use the chatbot through a CLI interface or integrate it into your applications. Here's how to get started:

  1. 🧩 API Integration: For developers, GridNLP offers API endpoints for seamless integration into your applications. Currently the api https://gridnlp.onrender.com/chatbot/?question= is live to be integrsted into your application.
  2. Custom Training: Modify the intents.json for your requirements and run the train.py in bot directory. The resultant model would be available in the models directory. Load them in the server.py for further applications

Installation

GridNLP build file is available in the repo for the recent updated apk file. For further development of the project or for personal use, follow the steps:

  1. The project directory contains the Flutter and Python dev envs combined. So for the app development, cd into the flutter/chatbot.
  2. For the python backend development, bot directory contains all the necessary files for training, adding new intents.json for custom training, simple cmd based chat etc.
  3. Run python -m venv .env for creating a new environment. Activate the environment and run pip -r requirements.txt to install the necessary packages.
  4. To run the server locally, simply do python server.py to start the FastAPI server. =======

Getting Started and Installation

GridNLP APK file is available in the release bundle of the repo. For setting up the server or re-configure the setup for your needs, follow the steps:

  1. Clone the repo and setup an environment in the directory by running python -m venv .env.
  2. Activate the environment
  3. Run pip install -r requirements.txt for installing the packages required for the project. The project was tested and developed for version 3.10>.
  4. Now, for custom training , cd to bots and change the intents.json for custom responses and model.py as per your need. After running train.py, the trained file would be present in the models directory.
  5. Make necessary changes in the server and run python server.py for starting the local server.
  6. For using the API, the server is hosted in render.com and the api for QnA could be done by using this :
    https://gridnlp.onrender.com/chatbot/?question=<your_query>
  7. For flutter development, cd into flutter/chatbot for accessing the flutter directory.

740981e0d3e30664abf875786a0f04a0c3c73476

Usage Examples

Here are a few examples of how GridNLP can assist users in substation maintenance:

  1. 🗣️ Query: "How do I maintain a transformer?"

    • Response: GridNLP provides step-by-step guidance on transformer maintenance, including safety precautions.
  2. 🗣️ Query: "What are the acceptable limits for surge arrester tests?"

    • Response: GridNLP explains the acceptable limits for surge arrester tests and how to conduct them.
  3. 🗣️ Query: "I encountered an issue during maintenance. What should I do?"

    • Response: GridNLP offers recommendations to troubleshoot and resolve common maintenance issues.

Industry Standards

GridNLP emphasizes compliance with industry standards and regulations. Some of the key standards it covers include:

  • 📊 IEEE Standards
  • 📜 ANSI Standards
  • 🔥 NFPA 70E
  • 🏭 OSHA Regulations

By adhering to these standards, users can ensure the safety and reliability of substation equipment.

Contributing

We welcome contributions from the community to enhance GridNLP's capabilities and accuracy. If you'd like to contribute, please follow our Contributing Guidelines.

License

GridNLP is released under the MIT License.

Support

<<<<<<< HEAD If you encounter any issues, have questions, or need assistance, please contact our support team at [email protected].

If you encounter any issues, have questions, or need assistance, please contact our support team at [email protected].

740981e0d3e30664abf875786a0f04a0c3c73476

GridNLP - Empowering Substation Asset Maintenance with NLP.