Project 'Control'

🚀 A software and hardware platform for studying Control Theory, Reinforcement Learning (RL), Shapers, and Modern Robotics.

---

📌 Overview

This project provides a unified environment for:

• Real-world robotics experiments
• Control Theory algorithms testing
• Reinforcement Learning (RL) training
• Hardware-in-the-Loop (HIL) testing

Supported platforms:

• Cart-pole (in work)
• Double pendulum (comming soon)
• Butterfly (comming soon)

---

🛠 Setup & Installation

Prerequisites

• Docker (Install Guide)

🚀 Quick Start with Docker

To open the documentation you need to run the following service

docker compose up --build get_docs

To run the test application, it is necessary to run the following service

docker compose up --build test

---

📂 Project Structure

├── docker-compose.yaml      # Docker Compose configuration for multi-container setup
├── Dockerfile               # Docker image build instructions
├── docs/                    # Project documentation (Sphinx)
├── firmware/                # Device firmware code
│   ├── arduino/             # Arduino-specific firmware
│   └── shared/              # Shared firmware components
├── LICENSE                  # Project license file
├── python_lib/              # Python library package
│   ├── docs/                # Library documentation
│   ├── setup.py             # Python package configuration
│   ├── src/                 # Source code
│   │   ├── inno_control/    # Main package
│   │   │   ├── core.py      # Core functionality
│   │   │   └── __init__.py  # Package initialization
│   │   └── main.py          # Main application
│   └── tests/               # Unit tests
├── README.md                # Project overview and documentation
└── requirements.txt         # Python dependencies

---

🛠 Development Guide

1. Local Development (Without Docker)

TO-DO

2. Development in Docker

Run docker container

docker compose up --build terminal

Attach to the docker from other terminals, by running:

docker compose exec terminal bash

3. Build Docker

After making changes to the dockerfile, you need to build a new image and upload it to the cloud

Build docker image

docker build -t image_name -f Dockerfile .

Tag docker image

docker tag image_name fabook/control:tag_name

Push docker image to Docker Hub (Optional)

docker push fabook/control:tag_name

Important

For reproducibility purposes, it is necessary to maintain the most current version of Docker on Docker Hub. But it is not necessary to call this command on every build.

Tip

The tag can be anything, but by default docker compose calls the image with the tag 'latest'

5. Git & Commit Rules

🔹 Core Tags (Commit Types)

Tag	Description
feat	A new feature. Example: feat: add user authentication
fix	A bug fix. Example: fix: resolve crash on null input
docs	Documentation changes. Example: docs: update README.md
style	Code style/formatting (e.g., spaces, commas). Example: style: format code per PEP8
refactor	Code refactoring (no functional changes). Example: refactor: optimize function X
perf	Performance improvements. Example: perf: reduce data loading time
test	Test-related changes. Example: test: add API test coverage
chore	Maintenance tasks (deps, configs). Example: chore: update webpack
ci	CI/CD changes (GitHub Actions, GitLab CI). Example: ci: add staging deploy
build	Build system changes. Example: build: add Dockerfile
revert	Reverts a previous commit. Example: revert: undo commit 123abc

🔹 Additional Rules

1. Message should be clear and concise.

   • ❌ Bad: fix: bug
   • ✅ Good: fix: prevent form submission error

2. Body (optional) — detailed description of changes.

   fix: fix memory leak in module X  
   
   The leak occurred due to unclosed DB connections during long sessions.  
   Added `cleanup()` to release resources properly.  
   

3. Footer (optional) — issue links, breaking changes.

   feat: add WebSocket support  
   
   BREAKING CHANGE: Legacy `/chat` API is deprecated.  
   Closes #123  
   

---

🤖 Hardware Setup

Connecting a Real Robot

1. Cart-pole

TO-DO

---

❓ Support & Contacts

• Robotics Lab: Room 105a
• Team contact: Evgenii Shlomov