League: RCJ Maze Byte Volta Software The robot's software system was designed to ensure precise and reliable motion control while navigating the RoboCup 2026 Rescue Country: Brazil Category : Maze Maze. To achieve this, four independent PID controllers (Proportional, Integral, and Derivative) were implemented, responsible for Brazil is a country known for its rich culture and growing emphasis on technology and innovation. In the Northeast maintaining wall alignment, correcting orientation angles, and ensuring region, the state of Ceará has become a rising hub for technological development, especially in the fields of directional stability. These controllers operate based on sensor fusion, robotics and automation. Ceará benefits from significant investments in digital infrastructure and a strong focus on which includes VL53L4CD time-of-flight sensors for accurate distance measurement, wheel encoders for velocity control, and filtering education, which together create an environment conducive to innovation. algorithms such as Kalman filtering and moving average to reduce Organizations like Farias Brito play an important role in preparing skilled professionals by combining theoretical sensor noise. This combination allows the robot to continuously adapt knowledge with hands-on experience in robotics and related technologies. This educational foundation helps fuel to environmental conditions, maintaining smooth and accurate PID Control Logic Diagram a vibrant local community of students, engineers, and entrepreneurs who are developing advanced robotic movement even in narrow passages or sharp turns. Font : Dirman Hanafi solutions to address real-world challenges. About us: Mapping and Navigation We are Byte Volta, the Brazilian team representing the educational institution Farias Brito. Our journey began in While moving through the maze, the robot performs real-time mapping using a matrix- October 2022. The name Byte Volta is a creative play on words in our native language, Portuguese: Byte sounds based representation to record wall locations and explored areas. To determine its path, like “bate” (which means "hit" or "collision") and Volta means "return" — together symbolizing both impact and it employs a hybrid approach combining DFS (Depth-First Search) and Dijkstra algorithms. DFS allows the robot to explore deeply into a path before backtracking, resilience. while Dijkstra provides a broader analysis, checking nearby paths and seeking shorter routes. This combined strategy ensures adaptive and complete navigation, enabling the Miguel A. Santos João V. G. Geiss robot to systematically explore the entire maze, avoid revisiting previously explored Eletrical/Hardware Eletrical regions, and return to the starting point using reconstructed paths. As a result, the system effectively handles various maze structures, including branches, dead ends, and Dijkstra Maze Exploration “I assembled the robot’s “I assembled the robot, prototyped complex routes, ensuring efficient and comprehensive exploration. Visualization mechanical components and the PCB, managed the hardware Font : Suhaib Al-Ansarry designed the circuitry for and components, and oversaw connecting all electronic parts and their integration into the system.” Navigation Algorithms: designed the robot in the CAD software.” João P. V. Silveira (C) Hardware/Software Kaleo S. Mendonça Software Hardware: ”I am responsible for both “I oversee the overall software The robot features a three-layer modular design that optimizes functionality and space utilization: hardware and software, especially architecture and the exploration in the areas of victim identification algorithms that drive the robot.” and parts of 3D modeling.” 1.Lower Layer 2.Middle Layer 3.Upper Layer (Base Platform): (Control Center): (Operational Interface): Optimized space allocation for upper Dedicated space for electrical Unobstructed camera field of view layers. components System status display interface Awards: Strategic component placement for electrical efficiency Suspension system. Efficient cable management Protection electronics. for sensitive Carrying handle for manual transport 2022 2024 1st place LARC, Latin American Robotics 1st place LARC (Rescue Maze) Competition (Rescue Maze) 3rd place LARI Competition Font : Authors, 2026 Font : Authors, 2026 Font : Authors, 2026 Best Interview Award LARC (Rescue Maze) 11th place OBR (Rescue Line) 4th place OBR (Rescue Line) 2023 2025 1st place LARC (Rescue Maze) Electrial Components: Innovation: Victim Identification To identify simulated victims throughout the maze, the robot features an embedded 20th place RoboCup (Rescue Maze) Participation Rescue Line The team’s Printed Circuit Board(PCB): The team designed a board, ilustred in the image, specially made thinking the bot computer vision system based on the OpenMV RT1062 camera. This system is capable of 4th place LARC (Rescue Maze) 16 th place RoboCup (Rescue Maze) 2nd place - knowledge fair functions, to comport our components, reduce the electrical structure, simplify and maximize the functionality also. Suspension and Wheels: real-time image processing using Convolutional Neural Networks (CNNs) for accurate The suspension system consists of two springs and operates pattern recognition. The software also includes multi-threshold color detection and circle as follows: the gearbox applies a contact force to the recognition algorithms, enabling reliable identification of visual targets and Greek silicone wheels, compressing the springs and lifting the letters. The system is optimized for performance, achieving a detection latency of under robot. Made of silicone, the wheels improve grip, and the 100 milliseconds and a classification accuracy of 98.5%, even in environments with Distance Sensor Color Sensor for Infrared Sensor for variable lighting. This capability ensures the robot can operate autonomously and combined system allows a maximum tilt angle of 13.5 Blue Tile Black tile degrees. efficiently, quickly and accurately identifying victims — an essential function in realistic Captured through our robot's ToF - rescue simulations. TCS34725 IT TCRT5000 vision system VL53L4CD Font : Authors, 2026 Computer Vision Acceleration and Change I2C Sensor Serial Monitor Display Turning Adress GY-BNO08X Display PCF8574N Nextion 2024 2025 Latin American Championship Latin American Championship Realease of the kits H-bridge Driver & Rescue Kit: Robot Evolution In 2024, we improved our PCB, Motor 63:1 Metal Controller The kit system features a servo motor that selects the side In 2023, we fixed the intermittent STS3032 - Gearmotor M5Stack - 4 for kit deployment. Strategically positioned ramps guide optimized our kit system — which Servo contact issue by implementing a 20Dx43L mm Encoder Driver the kit, ensuring accurate placement within the tile. This was too strong and caused the kits custom PCB, and addressed the 12V CB Module design improves reliability and precision during deliveries in to be pushed off the tile — and high current draw by replacing a competition. enhanced our computer vision by Font : author Raspberry Pi with two OpenMV implementing CNNs. boards. Camera Microcontroler PCB In 2026, for RoboCup, we In 2025, for CBR and RoboCup, we upgraded our suspension system, developed a suspension system for ESP32 P4 OpenMV RT1062 introduced new wheel sizes made the robot and new wheels made of with Ultrawide of silicone, improved our PCB, and silicone. Additionally, we improved Lens replaced the kit system for greater both our PCB and our kit system. precision.