🤖 Cleaning Robot Simulation (With and Without Memory)

This project simulates a robotic agent navigating a 5×5 grid to collect randomly placed leaf debris. The system is implemented in Python using Jupyter Notebook and ipywidgets for interactive visualization. The robot operates under two distinct modes:

• Without Memory: The agent cleans the environment without storing visited positions.
• With Memory: The agent keeps track of cleaned cells using a memory matrix and adapts its behavior to ensure full coverage.

---

🧠 Description

The robot starts with a defined amount of energy and a central starting position. As it navigates the grid, it attempts to remove leaf debris (🍂) scattered randomly on the board. It consumes one unit of energy per action (scan or move). The simulation ends when energy is exhausted or all debris is removed.

---

🔧 Technologies Used

• Python 3.9+
• Jupyter Notebook
• ipywidgets – for dynamic visual rendering of the grid
• NumPy – for efficient matrix management and logic

---

⚙️ How It Works

💡 General Behavior

• The board is displayed as a grid using HTML tables.
• The robot (🤖) moves in cardinal directions.
• It consumes energy while scanning or moving.
• Each cell may or may not contain a piece of debris (🍂).
• The robot removes a leaf if it occupies the same cell.

📌 Without Memory Mode

• The robot operates reactively.
• It only checks and cleans the current cell.
• Movement is random, without reference to past states.
• No internal representation of visited or cleaned cells is maintained.

📌 With Memory Mode

• The robot maintains a n × n memory matrix:
  • 0: Unvisited or uncertain
  • 1: Visited and clean
• In addition to scanning the current cell, the robot:
  • Scans its 4 neighboring cells (escaneoVecino)
  • Updates its memory based on findings
  • Attempts to follow a logical cleaning path
  • Executes a deep cleaning column-based strategy if no debris is found after several steps

---

🎮 Controls and Functions

Function	Description
llenadoAleatorio()	Randomly places leaf debris (🍂) on the grid
escaneo()	Scans and cleans the current cell
escaneoVecino()	Scans adjacent cells and updates memory matrix
avanzar()	Moves the robot one step in the direction of its angulo
girar(angulo)	Changes the robot’s orientation
revisionMemoria()	Checks whether all cells have been visited or verified
limpieza(agente, escenario)	Main cleaning loop (while energy and uncleaned cells remain)

---

🖼️ Visualization

The grid is rendered as a dynamic HTML table:

• 🤖: Cleaning robot
• 🍂: Leaf debris
• Empty cells: Clean and unoccupied

---

⚠️ Simulation Parameters

Parameter	Value
Grid Size	5 × 5
Initial Position	(2, 2)
Initial Energy	200
Debris Probability	50% per cell
Emoticons	🤖 (robot), 🍂 (leaf)

---

📊 Outputs

At the end of the simulation, the following are printed:

• ✅ Total leaves collected
• 🔋 Energy consumed
• 🧼 Whether the robot completed a full cleaning (with memory only)

---

👥 Credits

Josh Sebastián López Murcia