Desynchronizing the Digital Factory: Relational Intelligence at the Limits of Industrial Artificial Intelligence (AI)

This article examines how industrial artificial intelligence (AI) and digital twin technologies reorganize human labor in a high-tech electric vehicle factory in Vietnam. Corporate narratives portray digitally simulated factories as spaces where human and machine actions can be synchronized and managed through computational intelligence. However, drawing on interviews with welding workers and analysis of industrial AI discourse, I argue that what makes industrial AI work is the human work of “desynchronization.” Workers routinely move beyond digitally prescribed instructions, interpreting machine outputs and adapting to material contingencies that computational models cannot fully anticipate. To explain this process, I develop the concept of relational intelligence: workers’ capacity to coordinate between computational plans and changing workplace realities. Relational intelligence enables workers to complete tasks, accommodate uncertainty, and produce desired outcomes. Industrial AI therefore depends not only on machine intelligence but also on the undervalued human practices that sustain its operation. By foregrounding workers collaborating with industrial AI, who describe themselves as “robots that run on rice,” this article challenges techno-optimistic accounts of automation and digital twins. It argues that the intelligence sustaining industrial AI is not located in machines alone but emerges through workers’ relational intelligence in coordination with computational systems.