The human backend of smart shipyards: data stability and precarious work in the South Korean shipbuilding industry

The digitalization of manufacturing industries has generated extensive scholarship on job displacement and the invisible labour sustaining platform economies, yet the human work required to build data infrastructures within traditional industries remains underexamined. This paper examines the backend of smart shipyard initiatives in South Korea’s shipbuilding industry. Drawing on interviews with workers, managers, engineers, and industry experts at a major shipyard, we show that data stability – the consistency and reliability of data inputs necessary for digital twin operation – is not a technical parameter but an institutional outcome shaped by the shipyard’s three-tier employment structure. The secured core, the buffered middle, and the exposed periphery respond to backend labour demands through surveillance resistance, shallow compliance, and structural detachment, producing fragmented datasets that undermine the smart shipyard’s data foundation. We conceptualize this as the paradox of automation in segmented labour markets: automation depends on the very labour it seeks to displace, and dependence falls most heavily on those with the weakest incentives and least protection. Situated within the East Asian developmental welfare state, the study reframes data stability as a problem of institutionalized datafication, extending research on digital precarity beyond platform work to the backend labour of traditional industries.