CiberRato Robot Simulation Environment
Universidade de Aveiro / IEETA

Information

CiberRato Robot Simulation Environment simulates the movement of robots inside a labyrinth. Robots objective is to go from their starting position to beacon area and then return to their start position.

The MicroRato competition [http://microrato.ua.pt/], held annually at Aveiro University, uses these these tools for its Explorer league.

Contents

• simulator - The simulator source code
• Viewer - The Visualizer source code
• logplayer - The logplayer source code
• GUISample - Graphical robot agent (C++) source code
• robsample - robot agent (C) source code
• jClient - robot agent (Java) source code
• pClient - robot agent (Python) source code
• Labs - examples of labyrinths used in previous competitions
• startAll - script that runs the simulator, the visualizer and 5 GUISamples
• startSimViewer - script that runs the simulator and the Viewer

Additionally, this repository contains an agent developed for four competition challenges (see Challenges).

---

Installation

The source code was compiled with gcc/g++ (v9.3.0) using Qt libraries (v5.12.8) on Ubuntu 20.04.

1. Install dependencies

sudo apt-get install build-essential cmake qtmultimedia5-dev

2. Build the simulation environment

mkdir build
cd build
cmake ..
make

Start the Challenge-Specific Viewer

./startC1
./startC2
./startC3
./startC4

Running the Developed Agent

My developed agent can be found in the agent/ folder. It was used in four challenges of the MicroRato competition, where I achieved:

• Challenge 1 – 🥈 2nd place
• Challenge 2 – 🥇 1st place
• Challenge 3 – 🥈 2nd place
• Challenge 4 – 🥈 2nd place

Setup

1. Create and activate a Python virtual environment:

python3 -m venv venv
source venv/bin/activate   # Linux / macOS
venv\\Scripts\\activate      # Windows

2. Install Python requirements:

pip install -r requirements.txt

Run a Challenge:

Navigate to the corresponding challenge directory (C1, C2, C3, or C4) and run:

python main.py

⚠️ Important: You must be inside the specific challenge directory to start it. The challenges can be found at agent/C#/.

Challenges

Challenge 1 – Control

Objective: Control the robot’s movement through an unknown closed circuit as fast as possible, avoiding wall collisions.

Challenge 2 – Mapping

Objective: Explore an unknown maze and extract its map.

Challenge 3 – Planning

Objective: Explore an unknown maze, locate multiple target spots, and compute the shortest closed path visiting all spots, starting and ending at the starting position.

Challenge 4 – Localization, Mapping & Planning

Localization

• Navigate and localize the robot in an unknown maze using only noisy sensors (motors, compass, obstacle, beacon).
• No GPS available; collisions are penalized.

Mapping

• Explore the maze, extract the map, and identify all target spots.
• The agent must return to the starting spot after mapping.

Planning

• Compute a minimal-cost closed path visiting all target spots, starting and ending at the starting position.
• Noise parameters are specified in C4-config.xml.

---

Additional Resources

Presentations & Challenge Descriptions:

• challenges/presentation.pdf
• challenges/challgenge_c1_c2_c3.pdf
• challenges/challgenge_c4.pdf

Authors

• Nuno Lau, University of Aveiro, nunolau@ua.pt

• Artur C. Pereira, University of Aveiro, artur@ua.pt

• Andreia Melo, University of Aveiro, abmelo@criticalsoftware.com

• Antonio Neves, University of Aveiro, an@ua.pt

• Joao Figueiredo, University of Aveiro joao.figueiredo@ieeta.pt

• Miguel Rodrigues, University of Aveiro, miguel.rodrigues@ua.pt

• Eurico Pedrosa, University of Aveiro, efp@ua.pt

Copyright (C) 2001-2024 Universidade de Aveiro