Multi-Robot Firefighting System

This repository contains the implementation of a multi-robot firefighting system designed to extinguish fires in forest environments. The project leverages Artificial Potential Field (APF) controllers for navigation and coordination, ensuring collision-free operation and energy-aware task allocation among robots.

---

Features

• Navigation: Robots navigate through a forest environment using APF, avoiding obstacles while being attracted to fire zones.
• Task Allocation: Dynamic task allocation based on battery levels and proximity to fires.
• Energy Management: Robots monitor battery levels and return to a charging station when energy falls below a threshold.
• Simulations: Numerical simulations validate the system's efficiency and scalability.
• Barrier Certificates: Ensure collision-free movement using Control Barrier Functions (CBFs).

---

System Architecture

1. Robots:

   • Number: 5
   • Initialized at random positions.
   • Capabilities:
     • Extinguishing fires.
     • Avoiding obstacles.
     • Returning to a charging station when energy is low.

2. Environment:

   • Obstacles: Represented as circular zones (e.g., unburnt trees).
   • Fires: Fixed locations with decreasing intensity as robots work on extinguishing them.
   • Charging Station: A designated area where robots recharge.

3. Control Law:

   • Attractive forces guide robots toward goals (fires or charging station).
   • Repulsive forces avoid collisions with obstacles and other robots.

---

Mathematical Model

• Potential Functions:

  • Attractive: Pulls robots toward targets (fires or charging station).
  • Repulsive: Pushes robots away from obstacles and hazardous zones.

• Battery Management:

  • Battery decay rate: 0.025% per step.
  • Robots recharge upon reaching the charging station.

• Collision Avoidance:

  • Achieved through APF and Control Barrier Certificates.

---

Installation

Prerequisites

Ensure you have the following installed:

• Python 3.8+
• Libraries:
  • numpy
  • matplotlib
  • rps.robotarium
  • os

Installation Steps

1. Clone the repository:

   git clone https://github.com/pranaypalem/artificial-potential-field.git
   cd artificial-potential-field

2. Install dependencies:

   pip install -r requirements.txt

3. (Optional) Set up the environment for saving simulation frames: Update the frame_dir variable in the Python script to your preferred directory for saving simulation frames.

4. Download the fire image for visualization: Place an image named fire.png in the root directory of the project for simulation visualization.

---

Usage

Running the Simulation

To start the simulation:

python Final_code.py

Expected Outputs

• Simulation Frames: Captures of the robot movements.
• Trajectory Plots: Robots navigating toward fires while avoiding obstacles.
• Battery Annotations: Dynamic updates of battery levels over time.
• Fire Status Updates: Visualization of fire intensity reduction as robots work to extinguish them.

---

Key Functions

• attractive_force: Computes the attractive gradient pulling robots toward the goal.

• repulsive_force: Calculates the repulsive gradient for collision avoidance.

• within_fire_zone: Determines if a robot is within the radius of a fire zone.

---

Robotarium Simulation

The Robotarium platform was used to simulate and test the multi-robot system in a controlled environment. Key features of Robotarium usage include:

• Dynamic Testing: Verified collision-free navigation using barrier certificates.
• Visualization: Enabled real-time observation of robot paths and behaviors.
• Code Validation: Ensured the APF-based navigation and energy-aware strategies worked as intended in a physical-like simulation setup.

---

Simulation Results

1. Navigation:
   • Robots successfully reached fire zones while avoiding obstacles.
2. Energy Management:
   • Robots returned to the charging station when battery levels dropped below 25%.
3. Task Allocation:
   • Fires were extinguished sequentially, optimizing energy usage and efficiency.

---

Future Work

• Local Minima Handling: Improve APF to avoid trapping robots.
• Dynamic Task Allocation: Integrate learning-based methods for enhanced coordination.
• Real-World Deployment: Extend simulations to physical robots.