Recurrence and complexity analysis: from engineering signals to continuous improvement in manufacturing processes

Abstract

This work explores the application of recurrence analysis as a tool to evaluate complexity, stability, and predictability in engineering signals and manufacturing processes. Two case studies were implemented: (i) a motor under two conditions—without vibration and with induced vibration; and (ii) a simulated production process in two scenarios—without continuous improvement and with continuous improvement. The methodology consisted of generating time series, reconstructing trajectories in phase space, and computing recurrence plots along with their quantitative metrics. The results show that ordered and predictable systems (without vibration, with continuous improvement) are characterized by higher determinism and longer diagonal structures, while disordered systems (with vibration, without continuous improvement) exhibit lower determinism and shorter diagonals. These findings highlight the potential of this methodology to provide a quantitative perspective in both physical engineering systems and organizational processes, offering a valuable tool for diagnosis and decision-making in the context of manufacturing and continuous improvement.