ML-Keras Robotic Mechanism

A robotic control platform combining a real-time web dashboard, UDP socket networking, and AI-powered emotion detection. Built with accessibility at its core.

Live at: https://vermashaurya.github.io/ML-Keras-Bot

Overview

This project provides a full-stack interface for controlling a robot over a UDP socket connection. The web UI serves as the control surface — sending directional commands, streaming a live camera feed, running real-time emotion analysis, and supporting two-way chat. A Python tkinter GUI and UDP socket layer handle the backend communication.

Features

Web Dashboard

• Live robot status, quick D-pad controls, camera preview, emotion detection, and recent chat — all on one screen

Robot Controller

• Full directional control: Forward, Backward, Left, Right, Rotate Left/Right, Stop
• Keyboard shortcuts: Arrow keys, Q/E to rotate, Space to stop
• Timestamped command log

Camera & Emotion Detection

• Live webcam feed with REC indicator and timestamp overlay
• Real-time emotion analysis (Happy, Neutral, Sad, Surprised, Angry) with confidence bars

Chat

• Text and voice messaging with simulated responses
• Dashboard preview of recent messages

Python Backend

• PID/controller.py — tkinter GUI for local robot command dispatch
• Socket/server.py — UDP server on localhost:12371
• Socket/client.py — UDP client on localhost:12370

Project Structure

accessible-networking-ai/
├── Frontend/
│   └── UI/
│       ├── index.html       # Single-page app shell
│       ├── script.js        # All UI logic and state
│       └── styles.css       # CSS variables, theming, responsive layout
├── PID/
│   └── controller.py        # tkinter robot controller GUI
├── Socket/
│   ├── server.py            # UDP server (port 12371)
│   └── client.py            # UDP client (port 12370)
├── .github/
│   └── workflows/
│       └── deploy.yml       # GitHub Pages auto-deploy
├── .gitignore
├── LICENSE
└── README.md

Getting Started

Web UI

No build step required. Open Frontend/UI/index.html directly in a browser, or serve it with any static file server:

cd Frontend/UI
python3 -m http.server 8080

Then visit http://localhost:8080.

Python Backend

---

Requires Python 3 with tkinter (included in most standard installs).

Start the UDP server:

python3 Socket/server.py

Start the UDP client (in a separate terminal):

python3 Socket/client.py

Launch the tkinter controller GUI:

python3 PID/controller.py

Socket Ports

---

Component	Address
Server	localhost:12371
Client	localhost:12370

Keyboard Shortcuts

Key	Action
↑	Forward
↓	Backward
←	Left
→	Right
Q	Rotate Left
E	Rotate Right
Space	Stop

Shortcuts are active on the Dashboard and Robot Controller views.

---

License

Licensed under the Apache License 2.0.

Copyright (c) 2026