‌S Y S T E M ‌VERRIDE SO Robocup Junior Rescue Line EXPLORING NEW FRONTIERS IN THE FIELD OF ROBOTICS TEAM INDIA TEAM HARDWARE SOFTWARE We are System Override, a robotics team that started in 6th grade with Microcontrollers Line Following EV3 classes. Our early interest in technology turned into a commitment Our main microcontroller is the OpenMV H7 Plus mounted in a downward facing position to detect the line. It has direct We use the downward facing OpenMv camera for the lines, greens, silver tape control of the drive train motors. The bulk of the sensors and other actuators are controlled by a Raspberry Pi Pico 2W to designing and building autonomous robots. Together, we have and goal tiles. By using its inbuilt blob detection in 25+ ROIs, we are able to track which is subordinate to the OpenMV camera. We also have a secondary OpenMV camera mounted in a forward facing created reliable, autonomous, sustainable, open-source solutions that the positions and shape of all the visible line elements. We combine this position which acts as a sensor module for rescuing victims. information from all the elements into metrics such as slope and deviation to push the boundaries of robotics. Over the years, we have consistently Sensors calculate a steering value between -100 and 100 that dictates our bots movement demonstrated excellence on national and international stages, securing Due to the use of two cameras, our robot has reduced reliance on other sensors. This allows for greater hardware fault at any instant. Upon detection of greens we execute predetermined turns based on 1st place at RoboCup Nationals India (2026), 2nd place at RoboCup tolerance by virtue of simplicity. We have equipped our robot with a BNO055 9-axis IMU to accurately detect its the robots orientation followed by camera based realignment to the black line. orientation, 2 VL53L1X ToF distance sensors to detect obstacles and walls, and a TCS34725 RGB sensor to sense colors Nationals India (2023, 2024, 2025), and 3rd place at RoboCup on the ground. All the sensors communicate with the Pico subcontroller over I2C. Regionals India (2024). Internationally, we achieved 1st place in the Evacuation Zone Innovations Primary category at RCAP South Korea(2023), 1st place in the Since our base is entirely custom made we have worked on many ideas which were not possible with We enter the evacuation zone by detecting the silver tape using our downward Secondary category at RCAP China(2024), 1st place at the Singapore the retail kits we used in previous years. These include but are not limited to overlapping the wheel facing OpenMV camera. The RGB sensor is used to recheck silver tape to filter out and motor body to save space, custom silicone wheel treads that conform to the speed bump profile, false positives. Once in the rescue area, we open the arm and move around by Open (2025) and the Best TDP award at RCAP Abu Dhabi (2025). swapping gears in the DC motor gearbox to dial in the required torque/rpm profile, and greater modularity that supports selecting a random movement out of several predefined paths as the front facing easy assembly/disassembly and maintenance during the competition. camera looks for victims. Once a victim is found and picked up, we again use the LED Strip OpenMV front camera to search for its corresponding evacuation point and drop it. Once all Camera victims are rescued or the robot has exhausted its allotted rescue time, it searches N20 Motor for the exit by circling the perimeter. vl53l1X TOF Sensor Victim Detection Lipo cell We use the front facing OpenMV camera to identify victims in the evacuation zone through custom-trained machine learning models. After successfully entering the Ayaan Suri BnO055 rescue area, the camera scans for victims while the robot moves around. Once a Gyro Sensor Ayaan, our team lead, has been competing in RoboCup since victim is detected, the robot approaches, and the camera sends commands to the 2022. He has managed the hardware, electronics, and the Pico to pick up the victim using the claw. We also utilise embedded wires in the Pico 2W robot's controllers. He has trained the AI model and designed claw to confirm the presence of live victims. the robot in Fusion 360. Beyond technical work, he oversees the team’s documentation and helps manage its media presence. A I2C Multiplexer Tanash Garg CAPTURE READ GYRO FIND BLACK GA25 Motor START INITIALISE IMAGE VALUE LINE Tanash joined our team in 2025 and is our main programmer. He has helped train the model on Edge Impulse. Additionally, he STEER USING CALCULATE CORRECTION DETERMINE CORRECTION ACCORDING TO RAMP leads research on innovations for the robot, including exploring TCS34725 RGB STATE AND ERROR VALUE RAMP STATE VALUE and evaluating advanced sensors. He has also contributed to the SENSOR Lipo Battery team’s documentation and oversees its media presence. TB6612FNG MG90S ROBOTS JOURNEY Communication Motor driver LM2596 Buck Servo Motor GREEN DETECTED? True CHECK GREEN POSITION AND TURN ACCORDINGLY convertor For communication between the various controllers and sensors in our robot we use a False A combination of well established protocols. The main OpenMV and Pico use UART which allows for fast duplex data transfer. Most of the sensors in our robot communicate with the True GO BACK, ALLIGN MOVE FORWARD LINE LOST? TO CENTRE OF LINE UNTIL LINE IS A Pico using I2C over a multiplexer ensuring reliability of sensor data. The auxiliary OpenMV FOUND receives commands by the Pico over digital pin links. False 2022 2023 2024 2025 2026 Arm The arm is a simple yet effective design featuring 2 independently actuating claws to grab the victims and a sturdy frame to lift True We started with LEGO EV3. In our first year, we built a basic EV3 robot. In our them up. We have used N20 DC Motors for the claws as they are not as sensitive to external load and electrical noise as SILVER ENTER AND EXIT RESCUE DETECTED? RESCUE VICTTIMS STOP second year, we integrated Arduino with EV3 to help sort victims using Arduino Servo Motors while remaining just as compact. For lifting the arm we use a single MG995 servo motor that provides sufficient sensors. In our third year, we integrated EV3, Arduino, and an OpenMV torque and angular control for the task. True False camera, using the camera to detect victims and evacuation points. In our fourth Chasis year, we kept the same controllers but made the robot lighter and smaller by Our chassis is made of PLA and manufactured using an FDM 3D printer. After over 14 iterations we have come up with a RED OBSTACLE True CURVE AROUND ALIGN WITH DETECTED? removing the EV3 casing and adding a longer-lasting battery. In our fifth year, robust and modular design that utilizes a center mounted camera for line tracking and 4 GA25 6V motors in a tank drive DETECTED OBSTACLE LINE False we started building a fully open-source robot using an OpenMV camera for line configuration for locomotion. This chassis allows for uniform movement across many ramp and speedbump orientations following and a Raspberry Pi Pico 2 W for managing external sensors. simplifying the control algorithm. False