The "Luminosity Drone" project was undertaken as part of the E-Yantra Robotics Competition (E-YRC)
conducted by IIT Bombay. This year-long challenge pushed our team—comprising me(Team Lead), Sudhakar
Venkatachalam, and Chaitali Karekar—to develop a sophisticated autonomous drone system designed to
detect, localize, and map specific organisms within a controlled environment. The problem statement
required the drone to identify organisms marked with infrared (IR) LEDs and determine their global
coordinates with high precision.
To tackle this challenge, we focused on building a robust system that combined simulation, hardware
development, and advanced programming techniques. Our drone was equipped with both ROS 1 and ROS 2,
facilitating seamless control and communication across various components. The integration of ROS allowed
us to manage the complexity of the system, enabling efficient sensor data processing, motor control, and
inter-process communication.
Central to our approach was the implementation of a versatile search algorithm. This algorithm was designed
to ensure that the drone would thoroughly search the entire arena, regardless of its initial starting position,
and then return to the exact location for landing. With simple adjustments to a few variables, the algorithm
could be adapted to any map dimension, allowing the drone to search at desired altitudes and speeds.
The drone's control system relied on three PID loops to manage throttle, pitch, and roll, providing stable
and responsive flight control. All these functions, including the organism detection process, were executed
within a single codebase running at a frequency of 30 Hz, ensuring real-time performance and accuracy. For
organism detection, we utilized an onboard camera in conjunction with filters to identify the specific IR
LEDs representing the organisms. The camera fed real-time image data into computer vision algorithms, which
filtered out noise and accurately detected the organisms, enabling the drone to track and localize them within the arena.
Throughout the competition, our team faced numerous technical and logistical challenges. Developing a fully
autonomous drone required meticulous planning, extensive testing, and iterative problem-solving. The year-long
effort was both exhausting and exhilarating, pushing us to apply our knowledge in robotics, embedded systems,
and computer vision.
Unfortunately, a mishap during the final stages of the competition resulted in damage to our drone, preventing
us from submitting the last task. Despite this setback, we were proud of the progress we made and the skills we
developed along the way. Our work on the Luminosity Drone project not only placed us among the top 20 teams
in the competition but also provided us with invaluable hands-on experience in real-world robotics applications.
The Luminosity Drone project was an intense and rewarding experience that significantly enhanced our expertise
in robotics and autonomous systems. While we faced challenges and setbacks, the journey taught us the importance
of resilience, teamwork, and innovative problem-solving. The knowledge and skills we gained from this project
will undoubtedly serve as a strong foundation for our future endeavors in the field of robotics.