Industry 5.0 integrates industrial automation technology (IAT) with human-centric control systems to improve predictive maintenance, cost efficiency, operational reliability and production performance. However, traditional automation and conventional fuzzy decision-making (DM) approaches have several limitations in complex industrial environments. These methods often fail to represent vague, incomplete and uncertain information effectively and generally ignore the reliability and confidence level of expert judgments.
This paper develops an artificial intelligence (AI)-driven, multi-stage decision support system (DSS) for optimizing IAT within human-centric control systems. The proposed framework introduces the T-spherical fuzzy Z-number (T-SFZN), which combines the expressive capability of T-spherical fuzzy sets with the reliability structure of Z-numbers to handle uncertainty, hesitation, refusal and expert confidence simultaneously. Moreover, Schweizer–Sklar (SSK) t-norms and t-conorms are employed to construct flexible parametric operational laws for more adaptable information aggregation. The criteria importance through intercriteria correlation (CRITIC) method is used to determine objective criteria weights, while the multi-attributive border approximation area comparison (MABAC) technique is applied to rank feasible alternatives.
A comparative analysis with existing approaches confirms that the proposed T-SFZN-based DSS provides a more reliable, flexible, transparent and uncertainty-aware DM framework for Industry 5.0 applications.
The limitations of the above analysis stem from several factors that may impact the practical implementation of the proposed AI and machine learning (ML) solutions in IAT. First, the evaluation of alternatives relies heavily on generalized assumptions, which may not fully account for industry-specific variations, such as different production processes, regulatory requirements and infrastructure. Additionally, the application of T-SFZN and the Schweizer–Sklar aggregation operator (SSKAO) assumes that all data inputs are equally reliable and accurate, which may not always be the case in real-world scenarios where data quality can vary significantly. Furthermore, the complexities of AI systems, including their interpretability and potential biases, are oversimplified, and the solutions may not address unforeseen challenges like evolving cybersecurity threats or sudden market changes. The analysis also assumes that industries have the necessary resources to implement and maintain AI systems, which may not hold true for smaller or resource-constrained organizations. Lastly, the comparative analysis focuses primarily on the technical and operational aspects of the solutions, but does not sufficiently consider the human factor, such as resistance to change, workforce adaptability and the training required to manage advanced AI systems effectively. These limitations suggest that while the proposed solutions offer potential, further in-depth analysis and customization would be needed to ensure their successful adoption and long-term sustainability.
Industrial automation that uses AI and ML increases productivity and decreases reliance on manpower. Predictive maintenance is made possible by AI-driven solutions, which maximize equipment performance and save downtime. ML models reduce human interaction by enhancing production schedules, identifying irregularities and guaranteeing quality control. T-SFZNs, CRITIC and MABAC techniques are integrated to automate DM processes, producing more precise and reliable results. AI and employees can communicate seamlessly thanks to human-centric control mechanisms, which improve teamwork. As a result, there is less need for manual labor, expenses are reduced and total production efficiency is increased, resulting in a more economical and sustainable industrial environment.
By increasing resource efficiency and cutting waste, Industry 5.0's societal deployment of AI-driven technologies supports sustainable industrial practices. These technologies help create safer workplaces by eliminating the demand for manual labor and lowering health risks through the optimization of production performance and predictive maintenance. Fuzzy logic and DSSs work together to provide more precise and dependable operations that promote environmental sustainability and community well-being. AI's ability to optimize industrial processes also helps to create high-skilled jobs, boosting local economies and fostering a future in which businesses can strike a balance between social responsibility and technical growth.
The creation of an AI-driven, multi-stage DSS intended to maximize IATs within human-centric control systems is what makes this study unique. This research stands out by improving DM in complicated, imprecise situations through the creative integration of T-SFZNs to handle vagueness and uncertainty, as well as the application of SSK t-norms and t-conorms. Furthermore, a new way for assessing automation solutions is introduced by using the CRITIC method for weight determination and the MABAC methodology for ranking alternatives. The dependability and usefulness of these approaches are further shown by a comparison study.
