Advances in Adaptive Learning and Evolutionary Methods for Ground Robots

Autonomous ground robots have become fundamental platforms for a wide range of applications, including autonomous transportation, warehouse automation, environmental monitoring, precision agriculture, service robotics, and Human–Robot Interaction (HRI). As these systems are increasingly deployed in dynamic, uncertain, and human-centered environments, conventional control and navigation techniques often lack the adaptability, robustness, and autonomy required for real-world operation. In recent years, adaptive learning methods have significantly expanded the capabilities of robotic systems by enabling continuous improvement from data and experience. In parallel, evolutionary computation has emerged as a powerful paradigm for automatically designing controllers and behaviors, decision-making strategies. Combining evolutionary optimization with machine learning, reinforcement learning, symbolic methods, and model-based control is creating a new generation of hybrid intelligent robotic systems capable of adaptation, self-optimization, and lifelong learning.Despite considerable progress, developing reliable and scalable intelligence for autonomous ground vehicles (UGVs) remains an open research challenge. Real-world scenarios are characterized by uncertainty, dynamic obstacles, incomplete information, changing objectives, and frequent interaction with humans. Although adaptive learning techniques have demonstrated impressive capabilities, their practical deployment remains constrained by challenges related to sample efficiency, safety, explainability, sim-to-real transfer, and long-term generalization. Likewise, evolutionary methods provide effective mechanisms for discovering novel robot controllers, behaviors, and decision-making strategies. However, challenges remain in computational efficiency, scalability, online adaptation, and integration with modern learning paradigms.This Research Topic aims to bring together recent advances in adaptive learning and evolutionary methods for autonomous ground robots, with particular emphasis on hybrid intelligent systems that combine the complementary strengths of evolutionary computation, reinforcement learning, deep learning, symbolic reasoning, and classical control. The central objective is to improve autonomy, robustness, and adaptability in complex real-world environments.This Research Topic welcomes original research and review articles advancing the theory, methodology, and application of adaptive learning and evolutionary methods for autonomous ground robotic systems, particularly contributions that integrate multiple computational paradigms. Topics of interest include, but are not limited to:• Adaptive, lifelong, and online learning for robotic navigation, localization, path planning, control, and decision-making.• Evolutionary and hybrid intelligent methods integrating evolutionary computation, neuroevolution, reinforcement learning, deep learning, symbolic AI, and classical control.• Multi-objective optimization for behavioral adaptation, task allocation, and energy efficiency.• Sim-to-real transfer, domain adaptation, and safe, explainable, and trustworthy robotic AI.Applications include, but are not limited to, industrial automation, logistics, precision agriculture, healthcare, environmental monitoring, search and rescue, and autonomous transportation. Topic editors: Eddie Clemente, Eduardo Benitez Sandoval, Francisco Cruz, Jorge Fuentes-Pacheco, Marlen Meza-Sánchez, Hugo Estrada Esquivel Source: https://www.frontiersin.org/research-topics/85018/advances-in-adaptive-learning-and-evolutionary-methods-for-ground-robots