Algorithms and Methods for Designing and Scheduling Smart Manufacturing Systems Volume II

1. Introduction

Following the success of the first volume of this Special Issue, entitled “Algorithms and Methods for Designing and Scheduling Smart Manufacturing Systems” [1], we are pleased to introduce Volume II, entitled “Algorithms and Methods for Designing and Scheduling Smart Manufacturing Systems Volume II”. This continuation provides a new collection of high-quality contributions that reflect recent developments and trends in the field of smart manufacturing (SM). It addresses current challenges and proposes innovative methods for the design and production scheduling of next-generation manufacturing systems. As SM represents a significant paradigm shift in production systems and technologies, it is closely aligned with the principles of Industry 4.0 and related national strategies [2,3,4]. This aims to integrate advanced digital technologies (e.g., the Internet of Things (IoT), cyber–physical systems (CPSs), artificial intelligence (AI), robotics, cloud computing, etc.) into the manufacturing environment. The implementation of such technologies offers new possibilities for increasing efficiency, flexibility, and resilience in the face of global competition [5,6,7]. Moreover, the adoption of SM should not be seen merely as a technological upgrade as it also requires a thoughtful alignment with the established management philosophies and system design methodologies, including the alignment of recent advanced technologies. Recent research highlighted that lean manufacturing, reconfigurable, and cellular manufacturing concepts still play a significant role in the effective implementation of SM solutions [8]. Lean principles are crucial for designing flexible, efficient, and waste-optimized production processes in relation to the SM environment. According to Jian and Ding [9], the scheduling research needs to shift its attention and priority to real-time, smart scheduling and product-flow optimization using the latest AI tools. In this context, scheduling approaches based on reinforcement learning (RL) have proven to be useful [10,11,12,13]. RL is being increasingly applied to scheduling problems, enabling the development of intelligent agents that can learn optimal schedules through trial and error, interacting with an environment and receiving rewards for desired actions. This approach is particularly useful in dynamic and complex scheduling scenarios where traditional methods struggle to find optimal solutions. Explicitly, these approaches have been comprehensively mapped in review studies [14,15,16], which, among other things, have pointed out other current challenges in this area. These include the scheduling of production processes with integrated support for knowledge sharing and the growing demand for improvements in ubiquitous “intelligence” in production processes, including the design and implementation of intelligent algorithms.

Other important areas of research include modular and reconfigurable manufacturing systems, which provide the adaptability necessary to manage increasing and unpredictable demand in product customization. In addition, a reconfigurable manufacturing system can quickly adapt its structure and production configurations to meet new production [17]. In this context, the design and scheduling of manufacturing systems have become more complex and dynamic, thus requiring innovative algorithms and methods to address the latest industrial needs.

In recent years, research on SM has increasingly focused on human–machine collaboration, AI, and decentralized optimization. Hybrid human–AI decision-making approaches are explored to enhance trust and safety in cyber–physical manufacturing systems [18]. Current advances in multi-agent systems, genetic algorithms, and AI are also enabling real-time, distributed control without relying solely on centralized cloud infrastructure [19]. Scheduling algorithms are now more adaptive; they involve reinforcement learning (RL), multi-objective optimization, and stochastic modeling with the purpose of addressing uncertainties in demand, supply chains, and machine reliability [20,21]. System design is moving toward integrated platforms where simulation possibilities, digital twins, and related real-time data analytics integrate into unified frameworks [22].

This second volume of this Special Issue (SI) expands research in this area by presenting new theoretical solutions and practical approaches that contribute to the advancement of SM systems. This Special Issue contributes to both the theoretical and applied dimensions of SM by presenting a diverse set of case studies, optimization models, and diverse approaches that support the ongoing transformation of manufacturing systems toward intelligent, sustainable, and adaptive operations.

2. Description of the Papers

The presented Special Issue (SI) consists of six research articles presenting the latest works in the field. This Special Issue includes a diverse selection of articles that can be categorized into three thematic areas—(i) designing and scheduling manufacturing systems (four articles), (ii) sustainable production and packaging systems (one article), and (iii) digitalization and decision frameworks for smart manufacturing (one article). Most of the research contained in these articles focuses on solving complex industrial and logistic challenges through advanced optimization and decision-making techniques. These include genetic algorithms, hybrid metaheuristics, for example, the hybrid bat algorithm, and other intelligent approaches for system design and planning.

The abovementioned groups of articles are briefly described in the rest of this Editorial.

The paper entitled “Influence of Manufacturing Process Modularity on Lead Time Performances and Complexity” [23] written by V. Modrak, and Z. Soltysova investigates how modular process layouts influence two aspects of manufacturing systems, namely manufacturing lead time (MLT) and operational process complexity. This research included two case studies, where the authors applied the Tecnomatix simulation tool to investigate the impact of modularity on MLT and at the same time explore the relationship between structural modularity and operational complexity based on computational experiments. The obtained results showed a strong negative correlation between process modularity and manufacturing lead time as well as between process modularity and operational complexity. It confirms that manufacturing systems with higher process modularity can reduce time and simplify operations, and then modular design can provide benefits to enhance efficiency in the smart manufacturing domain.

The second paper written by G. Liu, M. Haung, and L. Chen, entitled “Optimization Method of Assembly Tolerance Types Based on Degree of Freedom” [24], proposes a novel optimization approach for the selection of appropriate geometric tolerance types in mechanical assemblies based on the degrees of freedom (DOFs) of tolerance zones. Traditional automated methods can often produce an excessive number of tolerance options, many of which may not align with geometric functional requirements. In connection with solving this problem, the authors present approaches based on control parameter degrees of freedom (CPDF) in the first phase of the research and subsequently use Boolean operations to map and compare degrees of freedom between tolerance zones and functional requirements. In conclusion, the algorithm proposed by the authors can effectively reduce the reliance on designer intuition and improve the scientific accuracy of tolerance selection, which was demonstrated by using an engineering case study to provide the validity of the method and its practical benefits.

The next article, entitled “Cell formation and intra-cell optimal machine location in CMS: a novel genetic algorithm (GA) based on machine encoding” [25], written by X. Wu, W. Li, M. Rizwan, Q. S. Khalid, M. Alkahtani, and F. M. Alqahtani proposed and validated a novel genetic algorithm for cell formation and machine layout in cellular manufacturing systems. This algorithm was tested using the MATLAB version R2022a simulation tool. The authors emphasized that the importance of manufacturing cell design and determining the optimal placement of machinery within the cell reduces material handling costs and improves efficiency. The proposed algorithm encodes machine locations and applies genetic algorithm operators to iteratively find better solutions. Their approach is tested using a real-world case from a Saudi Arabian automotive company and demonstrates improved productivity by optimized layouts.

The fourth manuscript bearing the title “Truck Transportation Scheduling for a New Transport Mode of Battery-Swapping Trucks in Open-Pit Mines” [26] is written by Y. Xiao, W. Zhou, B. Luan, K. Yang, and Y. Yang. The authors of this paper developed a scheduling model for truck transportation in open-pit mines. The aim of this proposed model is to minimize total haulage costs and truck waiting times by involving key operational constraints, including battery-swapping station availability, battery energy limitations, and the impact of ambient temperature on battery performance. The model integrates a range of real-world constraints and is solved using a basic genetic algorithm (GA) and an adaptive genetic algorithm (AGA). Comparative results show that the AGA significantly enhances scheduling efficiency by achieving a reduction in the total transportation costs, waiting times, battery swaps, and travel distances relative to the traditional GA. The proposed method holds the promise of achieving a more cost-effective and operationally reliable solution.

The fifth article, entitled “Decision Framework for Selecting Highly Sustainable Packaging Circular Model in Mass-Customized Packaging Industry” [27], written by authors R. Rajendran, and S. Ranjitharamasamy introduced a decision-making framework for selecting sustainable packaging models in mass-customized production. It provided a comprehensive decision-making framework for identifying the most sustainable circular packaging. This study leverages Z-number-based fuzzy approaches to address the inherent complexity, uncertainty, and subjectivity in sustainability evaluations. The proposed ZF-DEMATEL-TOPSIS method integrates expert judgment to assess and rank three circular packaging models (biodegradable, compostable, and recyclable) across five core sustainability dimensions broken down into fifteen specific criteria. Based on the analysis and obtained results, recyclable packaging emerges as the most sustainable choice. This framework offers small- and medium-sized enterprises a data-driven tool for making informed, sustainability-oriented packaging decisions aligned with the principles of the circular economy.

Last but not least, the article written by T. Raamets, K. Karjust, J. Majak and A. Hermaste, entitled “Implementing an AI-based Digital Twin Analysis System for Real-Time Decision Support in a Custom-Made Sportswear” [28], presents a real-world implementation of an artificial intelligence digital twin analysis system developed for small- and medium-sized enterprises (SMEs) specializing in custom manufacturing. Their proposed framework integrates real-time data collection, clustering-based analytics, and virtual simulation using the DIMUSA platform to support decision-making in human-centered manufacturing systems. This study demonstrates how production data, previously fragmented and manually managed, can be transformed into actionable insights by clustering techniques. Simulation methods were used to verify these insights and test scenarios. The authors emphasized in this paper that modular, scalable systems can facilitate the faster progress of the digital transformation of SMEs. In summary, the proposed study contributed to a gradually evolving methodology that integrates artificial intelligence and digital twins in the production environment of small- and medium-sized enterprises.

3. Conclusions

In conclusion to the second part of this Special Issue, we would like to note that editorial work is, on the one hand, an unknown story that motivates and encourages editors; on the other hand, its difficulty is often almost invisible and sometimes underestimated. In light of this, we hope that the contributions made to the second volume of this Special Issue will provide valuable insights to researchers, practitioners, and decision-makers involved in shaping the future of smart manufacturing in the digital era. To summarize the knowledge gained in this field, it can be said that future research will continue to be strongly influenced by the development of artificial intelligence, including all stages of production from product prototyping, production planning and scheduling, process designing, manufacturing, and post-manufacturing.