Adaptive Integrated AI Framework for Sustainable and Resilient Manufacturing Operations

Paper in proceeding, 2027

Artificial intelligence (AI) has become a key enabler of digital transformation in manufacturing, supporting applications such as predictive maintenance, process optimization, and quality control. However, these applications are predominantly developed in isolation, limiting their impact on system-level performance and sustainability. This paper takes the position that achieving sustainable and resilient manufacturing requires a shift from siloed AI applications toward adaptive integrated AI architectures. We propose a conceptual framework that connects production, maintenance, and quality domains through multimodal data fusion, cross-domain representation learning, and temporal modelling of system behaviour. These capabilities can enable adaptive optimization, where operational decisions are continuously refined based on system feedback and sustainability constraints. The proposed framework is applicable across diverse manufacturing environments, including discrete and continuous processes, and is motivated by real-industrial scenarios under an ongoing research project. This paper contributes by defining a new architectural perspective for AI in manufacturing and outlining a research agenda for the development of integrated, adaptive, and sustainability-oriented industrial systems.