IROS 2025

From October 19 to 25, 2025, the city of Hangzhou hosted the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2025), one of the world’s most important events in the field of robotics, bringing together more than 8,000 experts and researchers from around the globe.

Daniel Fuertes, from the Grupo de Tratamiento de Imágenes, presented the work entitled: “NaviFormer: A Deep Reinforcement Learning Transformer-like Model to Holistically Solve the Navigation Problem.”

This work introduces NaviFormer, a model based on deep reinforcement learning and Transformer-like architectures, designed to holistically and efficiently address the global navigation problem by combining both route planning (high-level) and trajectory planning (low-level).

Experimental results show that NaviFormer delivers high accuracy and computational speed, outperforming traditional approaches and demonstrating its potential for real-time missions.

We thank the IROS 2025 organizing committee for the opportunity to share our progress and exchange ideas with the international community. An inspiring experience that drives the future of intelligent robotics!