Experimental Design of Bio-Inspired Algorithms for Optimization Problems in Industry 5.0

Explainable AI (XAI) is a critical component of Industry 5.0, which focuses on human-centric, sustainable, and resilient manufacturing. Unlike Industry 4.0, which emphasizes automation and cyber-physical systems, Industry 5.0 integrates human intelligence with AI, ensuring that AI-driven decisions are transparent, interpretable, and aligned with human values. XAI enhances trust and collaboration between humans and AI systems by explaining AI decisions. This is particularly important in industries like manufacturing, healthcare, and finance, where AI-driven insights affect critical operations. With XAI, workers and decision-makers can validate AI recommendations, reduce bias, and ensure compliance with regulations like GDPR and AI ethics guidelines. 

The process parameters in casting, welding, renewable energy, and machinability research are optimized using mobile analysis of variance (ANOVA). Experiments are frequently designed using analysis of variance, which reduces the number of experiments for a given number of components and their levels. Compared to the Taguchi approach to design, it has numerous advantages. Experiments are carried out under the experimental design, and output and other reactions are documented. To determine which factors have a major impact on the answer, an analysis of variance is utilized. The process parameters are optimized to achieve a particular goal function, and regression equations are created to forecast the response. 

In the realm of wireless communication engineering, the localization of unidentified nodes in wireless sensor networks (WSN) is crucial. The localization of unknown nodes in wireless sensor networks (WSN) is an important problem in wireless communication engineering. Improving localization performance while extending the lifespan of energy-limited sensor nodes remains a key objective. The coordinates of unidentified nodes can typically be found using sensors that have the anchor nodes' known coordinates. However, it takes a lot of time to use bio-inspired algorithms. Therefore, it remains a difficult problem to find the best network configuration for node localization within a short period. This article describes an effective method for utilising an Elephant Swarm Water Search Algorithm (ESWSA) model to assess the ideal network parameters that lead to a low Average Localisation Error (ALE). The average localisation error (ALE) and the input variables’ transmission range (TR), node density (ND), anchor ratio (AR), and iterations (IT) are nonlinearly related in this study using an ANOVA-based model. In the experimental section, a metaheuristic optimization technique was applied to predict the ALE for a given set of input variables, producing an accuracy of about 97.843% and 98.109% for cross-validation and testing the dataset.

Modeling and optimization of the corresponding parameters of grinding processesare fascinating tasks for mechanical engineers. This paper mainly investigates a suitable model of the grinding process using a very popular statistical analysis tool, ANOVA, and Regression Analysis to control the output parameters, such as tangential force and change in temperature during grinding, by controlling the coolant flow rate, infeed of wheel, and scraperboard positioning. The experiment is also conducted to examine the effect of using a scraperboard on changing the grinding temperature. A Ktype thermocouple brazed beneath the workpiece surface and a force dynamometer are used for measuring the grinding temperature and tangential force during experimentation. Initially, a linear regression model is used to model force and change in temperature. Next, ANOVA is used for response optimization for the grinding process, where the objective is to minimize both output variables. From the optimization, it is found that the use of a scraper board is very effective in minimizing the tangential force and generated temperature during grinding.

Micro-powder injection molding (µ-PIM) has significantly evolved as a technique for manufacturing micro-components. It is well established that injection parameters play a critical role in achieving proper mold filling. Injection molding parameters significantly affect performance in terms of cost, product quality, and sustainability. Choosing the most effective (optimal) combination of injection parameters is therefore essential for each PIM procedure. To address this optimization challenge, a novel algorithm is employed in this study. An optimization technique based on ANOVA is used to predict good mold filling utilizing part weights. The proposed model takes injection pressure, temperature, and actual injection shot size as input variables and predicts the weight of the injected sample. The model is validated using the experimental results. The ANOVA-based Jaya Algorithm achieved maximum accuracies of 95.73% and 96.04% for the training and testing datasets, respectively. 

Electrochemical discharge machining (ECDM) performance enhancement is challenging when applied in the industrial field. The book chapter shows the secondorder correlating mathematical models with pulse on-time, electrolyte concentration, duty ratio, inter-electrode gap, and applied voltage for diametric over cut (DOC) and tool wear rate (TWR) when a mixture of potassium hydroxide (KOH) and sodium hydroxide (NaOH) is used as an electrolyte during blind hole generation on glass. ANOVA is used to analyze the fitness of models, as well as the effects of and process parameters. Desirability function analysis is performed in this chapter for the minimization of DOC and TWR for blind hole formation on glass and contour diagrams. Multi-objective optimization also represents the better results of micro-hole fabrication by the ECDM process. It is found that 50V/30mm IEG/20wt%/45 µs/.50 duty ratio/50Hz pulse frequency provides a minimum TWR of tungsten carbide tool and DOC of micro-hole. 

The augmentation of the characteristics of micro-ECDM during microdrilling on borosilicate glass is a challenging task. The book chapter indicates the ANOVA-based second-order mathematical models by correlating with voltage and conductivity of the electrolyte for radial over cut (ROC) and material removal rate (MRR) when NaOH is used as electrolyte during drilling on borosilicate glass. ANOVA is used to find the best fitness models for MRR and ROC. In this chapter, root mean square error (RMSE) for ROC and MRR is proposed during the drilling of glass by the ECDM process. It is found that optimized parametric combinations are 51 V/172 mmho-cm / 40mm, which achieve minimum ROC and MRR using stainless steel (SS) micro-tool.

Austenitic stainless steels are readily weldable, but sensitization and solidification cracks are common defects in the welding of such steels. It is found that the problem of solidification cracking remains under control if the weld pool solidifies in FA mode. 6-8% delta ferrite in the fusion zone can promote the FA mode of solidification. The effects of different semi-automatic MIG welding parameters, like welding current, gas flow rate, and welding speed, on solidification behavior and microstructure of the weldment are not clear. In this present work, an attempt has been made to optimize MIG welding input parameters to achieve 7% delta ferrite in the fusion zone. Initially, ANOVA-based analysis and optimization was performed. A new metaheuristic, namely the Elephant Swarm Water Search Algorithm (ESWSA), was used for optimization purposes to get the optimal condition of grinding.

Marine propulsion systems are evolving towards energy-efficient and environmentally friendly solutions, with solar-powered Switched Reluctance Motors (SRMs) emerging as a promising alternative. Despite SRM's inherent advantages, such as robust design, high efficiency, and fault tolerance, it suffers from torque ripples and non-linearity, which hinder its optimal performance in marine applications. This chapter investigates the performance enhancement of solar-fed SRMs in ship propulsion systems through the application of metaheuristic algorithms for intelligent control. Specifically, metaheuristic techniques such as Particle Swarm Optimization are employed to fine-tune controller parameters for minimizing torque ripple and improving speed control under varying load conditions. A solar-fed SEPIC converter is integrated into the system, providing a stable and renewable energy source while maintaining system’s efficiency. The proposed approach leverages metaheuristic algorithms to optimize the performance of the SRM drive by adjusting the control gains, thus ensuring smooth and efficient propulsion. Simulation results demonstrate a significant reduction in torque ripples, enhanced speed response, and improved overall system efficiency compared to traditional PI controllers. The findings highlight the potential of metaheuristic algorithms to transform SRM technology into a viable solution for sustainable marine propulsion systems. 

Advancements in wireless networks are crucial for the transformative potential of Industry 5.0, particularly in smart manufacturing and industrial automation. However, this new era demands solutions for persistent challenges, such as limited communication range, real-time data processing, and privacy concerns. Bio-inspired algorithms provide a novel approach to addressing these challenges, leveraging patterns observed in nature to enhance optimization and efficiency in wireless communication systems. This paper proposes an experimental framework using bio-inspired techniques to address specific industrial wireless communication needs, with a focus on edge intelligence, ultra-reliable low-latency communication, and energy efficiency. By emulating message protocols found in social animals, bio-inspired algorithms can optimize aspects like dynamic channel allocation and energy-saving topologies. The framework facilitates real-world testing in small to medium-sized industrial network settings, enabling an enhanced understanding of Industry 5.0 scenarios. Case studies demonstrate the effectiveness of bio-inspired solutions in optimizing network topology and energy consumption across smart factory networks, achieving notable results in scenarios requiring efficient data handling and reliable connectivity. This research highlights the importance of interdisciplinary approaches in supporting Industry 5.0 goals, demonstrating bio-inspired algorithms as promising tools for scalable, adaptive, and human-centered industrial wireless systems. The findings point towards sustainable advancements in wireless technology, fostering a deeper integration of nature-inspired solutions within next-generation industrial communication networks.

Recent developments in technology within the Indian banking sector, specifically the implementation of Artificial Intelligence (AI) and Robotic Process Automation (RPA), have significantly transformed banking operations and customer experience. This study examines the implementation strategies, applications, and benefits of AI and RPA in Indian banking by reviewing secondary data from government and private reports, publications, books, and journals. The findings indicate that AI is used for building knowledge graphs, recommending services, and detecting financial fraud, whereas RPA delivers AI insights in real time. Applications include customer service, risk management, fraud detection, and personalized financial advice. The benefits of integrating AI and RPA include operational cost reductions of 30-70%, improved accuracy, 24x7 operations, and enhanced customer experience. The challenges include data security, regulatory compliance, potential job shifting, and the requirement for staff training. The study concludes that the successful integration of AI and RPA requires a strategic approach that focuses on organizational readiness, stakeholder engagement, and ethical considerations. Future research should find the long-term effects, ethical implications, strategies for sustainable integration, and potential for promoting financial inclusion in India.

In a very fast communication world, the technology is inclining towards the 5G & 6G communication techniques. Among the available frequency bands for 5G, the n79 band (4.5 GHz) is very popular. It can be applied for many purposes like mobile communication, radios, etc. Even this band is also tested and deployed for 4G communication. A microstrip patch antenna is very popular for high-speed communication due to compact size, light weight, and ease of application. In this research work, we used Rogers 92 mL substrate, a copper ground and a copper patch with a hexagonal slot utilizing the inset feeding technique with the help of Computer Simulation Technology (CST) Studio Suite 2024. After adjusting the radius of the hexagonal slot and after shifting its centre to the exact location, satisfactory results have been achieved. For the best result, S11 falls up to -61.33dB at the exact 4.5GHz frequency and the corresponding bandwidth is 1.2919 GHz, which is suitable for 5G communication in the n79 band. Additionally, a mathematical model for the proposed patch antenna is developed using Artificial Neural Network models to predict antenna parameters like S11, bandwidth, and resonant frequency based on the radius of the hexagonal slot and the position of the centre of the slot. Predicted results show satisfactory accuracy in comparison with the CST Studio 2024 values.

This study focuses on optimizing a nonlinear liquid flow control process model using advanced techniques. A total of 134 datasets were collected, with 117 used for training and 17 for validation. To check the performance of the best algorithm, a comparative study has been conducted between GA, PSO, and the proposed hybrid GAPSO techniques. The performance was evaluated using MAE, MAPE, RMSE, and computational time. The results showed that the hybrid algorithm outperformed others in terms of convergence speed, computation time, and accuracy.

The conventional model is limited to obtaining the optimal solution from a nonlinear system objective function. An effective method for determining the optimal solution for nonlinear objective functions is the use of evolutionary algorithms (EA). In this study, we suggested three optimization techniques: Cuckoo Search Optimization (CSO), Particle Swarm Optimization (PSO), and a hybrid algorithm of CSO & PSO (HCSPSO) to determine the single and double-diode solar cells' optimal parameters. The proposed algorithm was compared with the CSO and PSO in terms of statistical metrics for a single-diode (SDM) and double-diode model (DDM). According to a comparison analysis, a more precise model for estimating a solar cell's optimal parameter with fewer iterations is offered by the upgraded CSO optimization. As a result, we suggested HCSPSO as the top optimization tool for delivering the most promising results. 

In this paper, a rectangular microstrip patch antenna with a triangular polygon at the top of the right side is designed. FR-4 is used as a substrate material with a relative permittivity of 4.3. The inset feeding technique is used to construct the proposed patch antenna. By changing three parameters, namely patch width (Wp ), substrate height (h), and width of feedline (Wf ), a dataset is created to achieve the resonant frequency of 26 GHz and a bandwidth of 2.5 GHz. ANOVA is employed to find out two equations for the resonant frequency (Fr ) and bandwidth (BW) to obtain target outcomes. All the simulations and calculations of the dataset are made using the CST Studio Suite version 2023, and ANOVA is carried out in Minitab version 21.