Comparative Analysis of Machine Learning and Deep Learning in Mobile Robot Development: A Systematic Review (#1811)

Authors

Ocaña Velásquez, Jesus Daniel

Castro García, José Heiner

Miranda Saldaña, Rodolfo Junior

Abstract

The rapid advancement of robotic technologies has driven the development of mobile robots, which are applied in various areas such as industry, logistics, exploration, security, healthcare, and many more. The aim of this research is to analyze and compare machine learning and deep learning techniques in the design and optimization of mobile robots with the purpose of identifying methods that stand out for their accuracy and reliability, thus contributing to the development of more efficient tools and models that increase the effectiveness of robotic systems in various environments. The PRISMA method was used to compile and systematize 65 articles relevant to the study topic. The results show that the development of mobile robots has become a frequent topic of interest for researchers in China and South Korea. In Machine Learning, the most prominent methods are Random Forest (RF) and SVM. In Deep Learning, the most outstanding techniques are CNN and SLAM. It is concluded that Machine Learning focuses on key applications such as navigation and mapping, while Deep Learning deals with complex challenges such as autonomous driving and image processing. Both disciplines complement each other in mobile robotics, where Machine Learning improves functionality and efficiency, and Deep Learning fosters innovation and understanding of the environment, opening opportunities for future research.