Digitization of End-To-End Value Streams: Conceptualizing a Data Model Based on Complex Graph Networks for AI-Based Tools

Value stream analysis and design are central elements of lean management, employed globally by industrial improvement teams to facilitate sustainable value creation. As value stream analysis and design evolve to incorporate material flow cost accounting, information logistics, and external influence factors, increasing data volumes and calculation complexity are driving up the expenses of method application. Traditional pen-and-paper approaches become impractical and non-value-adding. While research within the problem domain focuses primarily on analyzing current value streams, integrating real-time data, and facilitating ad-hoc optimization, the actual site of value creation—the target value stream design and corresponding improvement strategies for implementation—remains largely overlooked. To address this gap, this paper introduces a novel data model that translates value streams from physical whiteboards to the digital realm via complex value stream graph networks (CVSGN). This data model represents value stream structures as complex graph networks with nodes and edge feature vectors, providing a foundation for pattern-based digital value stream analysis and design through artificial intelligence-based tools. The data model offers industry practitioners practical benefits like optimized workflows, data-driven decision-making, and scalability, making it powerful for enhancing operational efficiency and supporting continuous improvement across diverse industries.