Digital innovation and sustainability transitions: Exploring the twin transition in energy-intensive processing industries

The thesis explores the emerging concept twin transitions (digitalization and decarbonization). It examines how digital innovation interacts with sustainability transitions by focusing on the case of the processing sector, which includes industries such as steel and chemicals. The thesis introduces two conceptual frameworks: (1) the Digital Meta-Regime, which captures the overarching influence of digital technologies across industries; and (2) Digital Innovation Systems, which provides a lens for analyzing how digital innovation operates within specific industries. Central to the research is the idea that while digital technologies like artificial intelligence (AI), digital twins, and advanced analytics hold transformative potential, their actual impact on sustainability is complex and contingent. A case study on AI in the steel industry illustrates this duality. Through an inventory analysis of 140 AI tools and 12 interviews with key actors, we demonstrate how AI facilitates operational efficiencies but tends to reinforce existing fossil-based paths rather than promoting novel low-carbon innovations. This reveals tensions where digital technology risks perpetuating unsustainable practices unless explicitly aligned with sustainability goals. However, the analysis also shows the wide-ranging applications that AI can enable within the steel industry like predicting process parameters; optimizing operations, scheduling, and electrical energy; and forecasting product demand, quality, and site emissions. Further qualitative analysis on the digital innovation system of the Dutch processing industry using 28 semi-structured interviews reveals key innovation resources, systemic problems, system-building activities, and higher-order mechanisms at play. For instance, various socio-technical barriers are faced by organizations involved with digital innovation, ranging from misaligned incentives, decision-making hurdles, and skills gaps to infrastructure, scaling, and partnership challenges, highlighting the challenges of integrating new technologies in industrial contexts. Understanding such digital innovation dynamics helped construct a policy framework for steering digital technologies towards sustainable applications, addressing shared barriers, and embedding better directionality. We argued for balancing critical perspectives with the enabling nature of digitalization, emphasizing the need for nuanced, context-specific policies to drive sustainability and prevent misuse of twin transition narratives by incumbents. To do so, we provide a heuristic tool to assist policymakers and industry actors in navigating twin transitions. The thesis provides theoretical contributions to the literature on technological innovation systems through the development and implementation of a resource-based perspective which provide new insights, beyond the traditional structure-function view. We demonstrate that variable accessibility and applicability are a quality of all technological systems, however, focusing on accessibility and applicability helps reveal unique characteristics within the innovation systems of pervasive or multi-purpose technologies. Additionally, the thesis provides groundwork on the co-dynamics of digital and sustainable innovation, the (ongoing) role of digitalization in shaping incumbent and niche socio-technical configurations, and how systems-thinking can inform twin transition policy. An important methodological contribution is made in combining and balancing technological (or techno-economic) and socio-institutional approaches for studying digital or other socio-technical systems. Further contributions and future work are discussed in relation to institutional logics, deep transitions, and mission-oriented innovation policy. Overall, this thesis provides important structure and guidance for thinking about digitalization in the context of sustainability.