The modern industries face a significant challenge regarding to the new product and resource evolutions and violate market demand. Reconfigurability is key, allowing the production of various product models within a part family. Companies, such as those in the automotive industry, adapt to product family evolutions taking into account future changes in technologies and resources, requiring frequent line reconfigurations. Reconfigurable Manufacturing Systems (RMS) offer a solution, enabling cost-effective and efficient evolution with product families. On the other hand, production resources and assets, such as robotic arms and conveyor belts, become prematurely obsolete, contradicting the principles of sustainability and the circular economy (CE). Utilizing second-hand resources for periodic production line changes presents an opportunity to promote circularity and efficiently respond to Industrial and supply chain disruptions while reducing costs and wastes. Moreover, we prioritize a human-centered approach to (up-)skilling assembly line workers to operate efficiently with second-hand resources, recognizing the importance of their roles in industry. Despite potential benefits from such a resource-sharing using CE, several obstacles hinder the adoption of circular practices, e.g., technical solutions for integrating second-hand resources, assessing the remaining lifespan of assets, and decision-making problems without compromising line productivity. Implementing such a CE in new generation of Industry with the uncertainty in demand and product requirements and manufacturing resource evolution is a critical issue for RMSs. Therefore, REALISTIC project aims to address these challenges by developing a tool to facilitate tactical decision-making processes related to resource selection, replacement, and assembly line design and reconfiguration. The primary goal is to provide an integrated tool that utilizes novel scenario-based mathematical programming models and efficient robust optimization techniques to support tactical decisions in future Industry. Predictive analytics techniques will be also developed to handle product evolution uncertainty, while robust optimization guides decision-making, ensuring insensitivity to worst-case scenarios. The main objective of the studied problems is focused on minimizing total costs related to resource selection, use, maintenance, addition, removal, relocation, and recycling for the worst scenario. By doing so, REALISTIC simplifies the design and future reconfigurations by developing optimization-based AI tools, benefiting industries economically and environmentally.