Characterization of Traffic Dynamics in Automated Guided Vehicle Systems: Analyzing and visualizing congestion patterns using causal chains and spectral clustering

Abstract

Automated Guided Vehicle (AGV) systems have revolutionized warehouse logistics by enhancing efficiency and reducing labor costs. However, the complexity of these systems can lead to traffic congestion, adversely affecting performance. This masters thesis investigates traffic dynamics in AGV systems using recursive data structures, interactive visualization, and spectral clustering; with the final aim being to describe and analyze congestion in these systems. Our study shows promise in streamlining design processes by providing more sophisticated analysis tools, data-structures for describing congestion, as well as novel ways to uncover and characterize large-scale traffic dynamics in the domain. Ultimately, this project could act as a stepping-stone for further research, by laying the ground-work for applying more formal methods.