State of the Trekking Autonomous Robots Competition

• Felipe G de M D'Elia USP
• Bruna Vitória R Pinheiro USP
• Larissa Driemeier USP

Resumo

This paper presents a comprehensive analysis of the 2025 RoboCore Experience's Trekking competition, one of Latin America's largest autonomous robotics challenges. Modeled after RoboMagellan, the Trekking category tasks ground robots with autonomously navigating a known environment, reaching predefined landmarks in sequence in the shortest possible time. We provide a benchmark of hardware and software approaches employed by participating teams, covering mechanical design, perception, planning, and control algorithms. Our survey reveals key trends, the dominance of Ackermann steering and differential drive systems, the widespread use of inertial navigation, and the growing adoption of ROS 2. Additionally, we examine team composition, highlighting the competition's relevance across high school, undergraduate, and graduate levels. The findings suggest that while simpler approaches remain competitive, advanced techniques such as sensor fusion and machine learning are gaining influence. We conclude with recommendations for future editions, emphasizing knowledge sharing, dynamic challenges, and ongoing benchmarking to reduce entry barriers and accelerate innovation in autonomous robotics.

Palavras-chave: Surveys, Technological innovation, Educational robots, Sensor fusion, Benchmark testing, Robot sensing systems, Market research, Planning, Vehicle dynamics, Autonomous robots, robotics competitions, navigation algorithms, RoboCore Experience, sensor fusion, educational robotics