The Sri Lankan Robotics Challenge (SLRC) is a distinguished event hosted by the Department of Electronic and Telecommunication Engineering at the University of Moratuwa. The competition is divided into two segments: the School Category and the University Category.
We are a group of five enthusiastic undergraduates from the Department of Electronic and Telecommunication Engineering at the University of Moratuwa, collectively known as "Sparkle." Our team members include:
- Pulindu Vidmal
- Akhila Prabodha
- Achira Hansindu
- Navini Jagoda
- Devnith Wijesinghe
Our team secured a commendable top 15 rank in the university category and reached the finals of the competition.
Our project was inspired by the theme of infinity stones from the Avengers movie series. For a detailed explanation of our tasks, please refer to the attached PDF document.
Our tasks were segmented into three thematic regions, each inspired by different celestial settings:
- Planet A - Mountains of Vormir
- Planet B - Ruins of Sakaar
- Planet C - Thanos’s Home
We implemented an IR array combined with a PID controller to ensure accurate line tracking. The IR sensors detect the contrast between the white line and the black surface, allowing the PID controller to make real-time adjustments to the robot's path.
Using OpenCV and computer vision with a web camera, we identified wall colors. This system helps maintain a consistent distance from the walls while detecting necessary colors for navigation.
Strategically placed under-mounted color sensors were used to detect color junctions. These sensors are calibrated to recognize specific color patterns at junctions, guiding the robot through turns accurately.
We utilized TensorFlow and OpenCV to identify object shapes, distinguishing between cylinders and cubes. This technology combination provides robust shape recognition capabilities.
Conductivity testing was employed to identify metal boxes, and a robotic arm was designed for precise grabbing. The conductivity sensor differentiates between metal and non-metal boxes, while the robotic arm handles pickup and placement tasks efficiently.
Three ToF sensors were used to measure obstacle heights accurately. This data helps the robot calculate the total number of gems collected based on obstacle heights.
Each challenge was tackled using sophisticated sensors and control algorithms:
- Line Following: Employed IR array and PID controller for precise path adjustment.
- Wall Color Detection: Used OpenCV with computer vision from a web camera to identify wall colors.
- Color Junction Detection: Implemented under-mounted color sensors for precise junction recognition.
- 3D Object Detection: Used TensorFlow and OpenCV to identify shapes, distinguishing between cylinders and cubes.
- Metal Box Detection and Grabbing: Conductivity testing and robotic arm for efficient handling.
- Obstacle Height Detection: ToF and ultrasonic sensors for precise height measurement.