Integration of YOLO (You Only Look Once) AI and Lean methodology for safer and more efficient work environments. A systematic literature (#786)

Authors

Guizado Atanacio, Grace Ariana

Meza Quispe, Andrea Fiorela

Canahua Apaza, Nohemy Miriam

Abstract

The integration of computer vision technologies with continuous improvement methodologies has become a key strategy for optimizing safety and efficiency in industrial environments. This study analyzes the combined impact of the YOLO algorithm and the Lean methodology on monitoring the use of PPE and improving operational performance in the manufacturing sector. A systematic methodology based on the PICO and PRISMA guidelines was used, covering publications between 2020 and 2025 extracted from databases such as Scopus, IEEE Xplore, and Web of Science. After a rigorous selection process, 104 relevant studies were analyzed, allowing for the identification of trends, applications, and knowledge gaps. The results show that YOLO, especially in its most recent versions such as YOLOv7, has achieved accuracies of over 95% in real-time PPE detection, surpassing traditional methods such as RFID sensors or classic computer vision algorithms. At the same time, Lean tools such as 5S+1, Kaizen, and Poka-Yoke have proven effective in reducing errors, standardizing processes, and promoting a preventive culture. The combination of both approaches enables more accurate automated monitoring, improves regulatory compliance traceability, and reduces risk exposure. It is concluded that the synergy between YOLO and Lean constitutes an innovative and effective solution for strengthening workplace safety and operational efficiency. However, limitations related to empirical validation in real environments and technological adoption in small businesses are identified, which presents opportunities for future research aimed at scalable and sustainable implementation.