Title: From Summer Learning to Swarm Intelligence: Robotics Club IIT Ropar at Inter-IIT Tech Meet 13.0
Introduction
The Inter-IIT Tech Meet has long been a stage for some of the brightest student minds in the country to demonstrate innovative thinking and engineering prowess. At the 13.0 edition hosted by IIT Bombay, the Robotics Club of IIT Ropar took on the Bharat Forge Problem Statement (PS), and in doing so, embarked on a journey that transformed not only their skills but also the club's technical foundation. This blog revisits the incredible effort by our previous core team, who represented IIT Ropar and built a full-fledged swarm intelligence system from scratch.
The Challenge: Centralized Intelligence for Swarm Navigation
The Bharat Forge PS tasked teams with building a "Centralized Intelligence System for Dynamic Swarm Navigation." The goal? To enable a swarm of robots to autonomously explore, map, and navigate dynamic, GPS-denied environments—all without any manual intervention. This meant designing an end-to-end multi-robot system capable of SLAM (Simultaneous Localization and Mapping), autonomous decision-making, and collaborative task execution.
Where It Began: Summer and ROS2
The journey began in the summer, when the core team decided to push beyond their comfort zone and pick up a completely new tool: the Robot Operating System (ROS2). What started as a learning initiative quickly became the foundation for the club's entry into the Inter-IIT Tech Meet. The team immerses itself in simulation environments, message-passing systems, path planning algorithms, and the intricacies of ROS2-based control architectures.
Team Members (Contributors from IIT Ropar):
System Architecture: Step-by-Step Development
The project was tackled in two major parts: map exploration and task allocation.
1. SLAM-Based Mapping
The team began with basic map generation using LiDAR-based SLAM. Utilizing gMapping and PLICP algorithms, they explored environments manually using teleoperation to better understand the baseline mechanics of mapping.
2. Autonomous Exploration (Single Bot)
They then implemented Frontier-Based Search (FBS) and RRT* algorithms to allow a single robot to autonomously explore unknown spaces. FBS, in particular, became the backbone due to its efficiency and lower computational overhead.
3. Multi-Robot Deployment
The next challenge was scaling the solution to multiple bots. Each robot operated in its own namespace, simulated in Gazebo, and performed SLAM independently.
4. Map Merging
A global map was created by merging local occupancy grids from each robot using transformation matrices and probability-based methods. This ensured coherence across multiple perspectives.
5. Task Allocation Engine
The system used a lightweight custom solution with an SQLite3 database to track pending and completed tasks. To assign tasks optimally, the Hungarian Algorithm was implemented to minimize overall traversal distance.
6. Scalability and Real-Time Control
The final architecture supported real-time obstacle avoidance, task reassignment, and dynamic scalability for any number of robots.
Key Tools and Technologies:
Impact and Legacy
The work done by the previous core team was not just a submission; it was a stepping stone for the club's future. The complete codebase, technical presentation, and research paper have been archived and shared openly to assist future teams.
Explore the Work:
Access the full GitHub repository containing the code, presentation, and research documentation:
https://lnkd.in/dkPTczxr
Conclusion
This project exemplified what a passionate and curious team can achieve with determination and the right learning mindset. From learning ROS2 in the summer to representing IIT Ropar at a national-level technical meet, the team showed what it truly means to learn by building. Their efforts continue to inspire upcoming batches and set a new technical benchmark for the club.
#IITRopar #RoboticsClub #InterIIT #ROS2 #SwarmRobotics #AutonomousSystems #BharatForgePS #TechMeet13 #LearningByDoing #OpenSource #Legacy