ABU Robocon 2022 - R2

The R2 robot was designed to excel in the dynamic and challenging tasks required by the ABU Robocon 2022 competition. This robot combined precision mechanics, high-speed control systems, and advanced sensors to perform in the high-pressure environment of the event.

Our Journey for the 2022 Season

Watch this video to see the journey and efforts of our team during the 2022 Robocon season, showcasing the dedication and hard work behind our robots.

My Role and Contributions

The robot was developed over almost a year, starting in September 2021, with the competition held in July 2022. In 2021, I was a senior member of the team and went on to lead the Embedded Systems and Software team in 2022.

Our first goal in August 2021 was to publicize the team and generate excitement about robotics. We achieved this by hosting a Robotics Workshop, where we taught 120 students how to build their own IoT-enabled line-following robots, using a kit developed by my team and me.

A recruitment drive was then held to bring new members into the team. I recruited 20 members under me, with the number narrowing down to 6 by the end of October. I mentored and taught them Embedded C, Arduino, MATLAB, and debugging concepts, along with topics like sensor fusion, path planning, obstacle avoidance, and best practices using Git.

While helping the new recruits get up to speed, the mechanical team began designing the robot according to the competition's problem statement. I started testing various control methods, microcontrollers, algorithms, and also took on side projects to help fund the team.

Additionally, I developed low-level code modules to enable us to focus on higher-level software later. These modules were designed to interact with every hardware component available to the team, including RoboClaw Motor Drivers, Benewake LiDAR sensors (TF02, TF03, TF Mini, TF Luna), SICK laser distance sensors, VNH2SP30 motor drivers, custom in-house motor control boards, pneumatic controls, gyroscopes, and more.

Autonomous Drivetrain Development

As the competition grew closer, I started developing autonomous software for the drivetrain.

I tested various paths for the competition using a cascaded PID controller to control acceleration, velocity, and position of the drivetrain in three degrees of freedom (X, Y, and Yaw). I used a quadratic function to calculate the optimal path for the robot, as seen in the video.

I faced issues like an uneven surface, causing the LiDAR sensors to detect above the arena fence, giving false distance values. Another challenge was drift from the Inertial Measurement Unit (IMU). I solved these issues by filtering out spikes in the data over a large interval for the laser measurements and using two IMUs mounted on damping materials in opposite orientations. This configuration helped balance out the drift, achieving a resultant drift of less than 1 degree over 5 minutes.