Robotics

For my final project in Robotics, I explored swarm robotics by developing and testing a multi-robot foraging system using the ARGoS simulation platform. I accomplished this work with the help of my project partner, Sanniva Halder Saha. The study examined how groups of foot-bots cooperate to gather resources under varying conditions—both in obstacle-rich and obstacle-free environments. This work not only deepened my understanding of robot behavior and coordination but also highlighted challenges in scalability, robustness, and energy efficiency.

Objectives

• Demonstrate Collective Behavior: Investigate how simple individual rules lead to complex swarm behavior.
• Evaluate Environmental Impact: Compare performance differences in environments with and without obstacles.
• Optimize Foraging Strategies: Analyze state transitions (exploration vs. resting) and energy management to improve resource collection.

Methodology

The project was implemented using ARGoS 3.0, a state-of-the-art simulation tool for large-scale multi-robot experiments. Key elements include:

• Robot Configuration: The foot-bots were equipped with differential steering actuators, LEDs for communication, range and bearing sensors, proximity sensors, and more. These sensors enabled the robots to navigate, avoid obstacles, and communicate within the swarm.
• Simulation Setup: Experiments were conducted in a 15m x 15m x 6m arena with 100 resource items scattered throughout. Simulations ran for 900 seconds with varying numbers of robots (ranging from as few as 50 to as many as 500) to test different swarm sizes.
• Experimental Scenarios: Two primary scenarios were evaluated:

  • Obstacle-Free Environment: Allowed observation of pure foraging efficiency.

  • Obstacle-Populated Environment: Introduced real-world challenges such as navigation and collision avoidance.

Data Analysis & Key Findings

• Behavioral Dynamics:

  • In large swarms, a high initial rate of movement (exploring bots) was observed followed by a quick transition to resting states. This was more pronounced in the presence of obstacles.

  • Optimized control parameters led to a more balanced distribution between walking and resting states, enhancing overall efficiency.

• Foraging Efficiency:

  • Experiments demonstrated that mid-sized swarms and the optimized setup collected resources more consistently than very small or very large groups.

  • The presence of obstacles slightly reduced foraging rates due to congestion and navigation challenges.

• Energy Management:

  • Managing energy consumption was crucial. The optimized version showed a slower energy decline, suggesting better energy efficiency even when handling complex environments.

• Scalability & Adaptability:

  • While the swarm demonstrated strong self-organizing behavior, scalability issues emerged with larger groups. The insights gained from these challenges have paved the way for future improvements in control strategies and inter-robot communication.

This experimental study provided valuable insights into the dynamics of swarm robotics. The results underscore the importance of balancing exploration and rest, efficient energy management, and adaptive behavior in dynamic environments.

Technical Details & Resources

• Simulation Platform: ARGoS 3.0.0 beta59 running on MacOS Sonoma 14.

• Programming: Robot controllers were developed in C++.

• Experiment Files & Setup: Detailed instructions and simulation files are available on GitHub.