Journal of Advances in Industrial Engineering
Volume:58 Issue: 1, Winter and Spring 2024

Both man-made and natural disasters can cause significant damage to property and human lives. Giving emergency medical services to the casualties as fast as possible after a disaster is critical. However, the destruction of some infrastructure such as roads, in the aftermath of a disaster, makes this process complicated. Artificial intelligence is now more frequently used to solve a wide range of difficult problems. In this paper, a combination of a deep learning model and particle swarm optimization algorithm is proposed to extract roads from satellite images, which can be useful for emergency vehicle drivers to recognize the best available path to reach casualties in disaster zones and give medical services to them faster. The model is evaluated by the evaluation metrics. Moreover, it is compared with other common models. The proposed model shows remarkable performance and 92% accuracy. Also, some predictions based on the model will be presented.

Scheduling for flexible flowshop environments is generally limited by resources such as manpower and machines. However, the majority of efforts tackle machines as the only constrained resource. This paper aims to investigate the problem of scheduling in flexible flowshop environments considering different skills as human resource constraints to minimize the total completion time. In this way, a mathematical model of complex integer linear programming is presented for solving small-sized problems in a reasonable computational time. In addition, due to the NP-hard nature of the problem, a whale hybrid optimization algorithm is tuned to solve the problem in large-sized dimensions. In order to evaluate the performance of the proposed optimization algorithm, the results are compared with five known optimization algorithms in the research background. All evaluations and results show the good performance of the whale hybrid algorithm. Especially, the final solution of the proposed algorithm shows a 0.75% deviation of the best solution in solving different instances on large-scale sizes. However, the genetic algorithm, memetic global and local search algorithm, and hybrid salp swarm algorithm are in the next ranks with 3.31, 3.52, and 4.02 percent respectively. In addition, proper discussions and managerial insights are provided for the relevant managers.

Product configuration plays a vital role in product customization. Customers require products with reasonable cost and reliability, so manufacturers should exchange between reliability and production cost through product configuration. To investigate this problem, a novel combined FTA-DFMEA method is presented that implements integrated AHP-TOPSIS to improve product configuration. In this procedure, customer’s needs and market’s feedbacks are considered to identify possible product failures, and an integrated AHP-TOPSIS is applied in order to select the most crucial potential failure based on some identified and extracted criteria. Then, minimal paths are obtained through fault tree analysis and an inverse search method is done to identify related functions and defective components. Failure modes and effect analysis is implemented to conclude modes of failure, effects, and causes. Subsequently, a combined AHP-TOPSIS method is utilized for ranking failure modes and selecting the most crucial failure mode. Failure modes are addressed according to their importance and corrective actions are carried out to improve product configuration. Suppliers with various policies, reliability, warranty and purchasing costs are considered. In addition, for the first time all configuration models like series, parallel, and joint series-parallel as well as redundancy allocation are taken into consideration. A minimum improvement index is considered, which is determined by the decision-maker based on risk-averseness. Eventually, a case study of a laptop system is introduced to evaluate the practicality of the developed algorithm. The results indicate that the proposed method creates different efficient alternatives for the decision-maker to enhance reliability, total costs, and product configuration. Also, the proposed framework consisting of the integration of failure analysis and MADM techniques, effectively identifies failure modes and prevents them from occurring.

Today's food supply chains are increasingly vulnerable to uncertainties in both supply and demand, as well as unexpected disruptions. Broiler supply chains, among the most vital globally, are no exception. To address this, the proposed model for this essential product, spanning five tiers and covering 31 states, incorporates resilience strategies such as backup facilities and multiple sourcing. The model utilizes bi-objective, multi-period, and multi-product mixed-integer linear programming to account for all three pillars of sustainability. The primary objectives are to maximize total supply chain profit while minimizing carbon dioxide (CO₂) emissions from transportation. Real-world deterministic data is imported into the model, which is solved using General Algebraic Modeling System (GAMS) software. The ε-constraint method is employed to generate Pareto-optimal solutions for the competing objectives. Additionally, validation and sensitivity analysis are conducted on key parameters within reasonable ranges. The results demonstrate an enhanced network that is both more profitable and less environmentally harmful.

The healthcare sector has recently encountered significant challenges, including limited funding and intense competition. These issues have adversely impacted hospital supply chains, resulting in budget cuts, staffing shortages, and logistical difficulties. This study introduces a novel two-step clustering approach to address the multi-depot vehicle routing problem (MDVRP) in healthcare logistics, specifically focusing on optimizing the delivery of pharmaceutical supplies to hospital pharmacies in Tehran. The method begins with the K-means algorithm to identify optimal distribution centers in the first step. In the second step, K-means clustering, incorporating vehicle capacity and demand values, is applied to each distribution center to allocate demand points for each vehicle. The vehicle routes are then determined by solving the traveling salesman problem. By optimizing the number of distribution centers using the silhouette score, which resulted in a score of 0.3567 for four centers, the study shows that deploying five vehicles from four strategically located centers can meet the needs of Tehran hospitals with a total travel distance of 119.68 km. A comparative analysis with two alternative methods reveals that the proposed approach offers a 14% improvement in minimizing the total travel distance. This method not only helps identify optimal locations for new distribution centers but also develops efficient routing plans for pharmaceutical distribution, ultimately reducing costs and improving service quality within healthcare logistics.

Control charts and maintenance strategies are essential tools in production management. However, despite their inherent connection, these tools are often studied and applied independently. To better reflect real-world scenarios, this paper focuses on the economic design of an integrated production planning model based on a synthetic adaptive Exponentially Weighted Moving Average (EWMA) control chart. To mitigate machine failure rates, two types of maintenance strategies are incorporated: reactive maintenance (RM) and preventive maintenance (PM). The model uses the particle swarm optimization (PSO) metaheuristic algorithm to minimize the total production cycle cost while adhering to statistical quality constraints. A comparative analysis is conducted to evaluate the impact of variable sampling intervals in control charts on overall costs. Sensitivity analysis is performed to examine how model parameters influence optimization policies. Finally, the results are compared with previous studies to demonstrate the effectiveness of the proposed method.

Nurses’ scheduling problems have attracted a significant amount of healthcare research, indicating the importance of these issues. In this paper, it has tried to present a multi-objective model for the assignment of nurses and anesthesiologists to surgical teams, considering frequency and fairness in allocating time to staff members. Since idle time is inevitable, we seek to divide idle time equally among staff. In addition, the break time of each staff member have an almost regular frequency during the shift. Minimizing overtime costs and maximizing attention to the willingness of surgical staff members to work overtime are other objectives of the problem. Three metaheuristic algorithms NSGA-II, MOPSO, and SPEA-II used to solve the presented model. A hybrid multi-objective genetic algorithm based on variable neighborhood search is also presented. The comparison of the solutions of 4 algorithms shows that the proposed hybrid algorithm has a significant superiority compared to other algorithms in terms of the average value of the solution, the quality of the Pareto solution set, and execution time. The presented model is compared with the real data of the surgical department of elective patients of a government hospital in Qazvin province. The obtained results show that the presented model has significantly created equality in the amount of working time of nurses and anesthesiologists in the elective surgery department. It has also spread the idle time of each staff member during the work shift, which has caused different time breaks for each one.

In today's competitive business environment, particularly for small and medium-sized enterprises (SMEs) in the information and communication technology (ICT) sector, survival hinges on the ability to innovate. Business Model Innovation (BMI) plays a pivotal role in driving competitiveness and enhancing performance. However, many organizations face organizational inertia, which can impede their ability to adapt to environmental changes, limiting their capacity to evolve. Organizational competencies and capabilities may play a crucial role in how businesses respond to these challenges. This study investigates the relationship between organizational inertia and BMI, focusing on the moderating effects of open innovation competencies and Information Technology (IT) ambidexterity. The research targets software companies in Tehran Province, Iran, within the ICT industry. The findings reveal that organizational inertia has a significant negative effect on BMI, highlighting the need to address inertia to foster business model innovation. Additionally, open innovation competencies serve as a key moderator in this relationship, suggesting the importance of developing these skills within organizations. Conversely, the moderating effect of IT ambidexterity was found to be insignificant.

This paper addresses the need to investigate customer behavior and incorporate competition, given the significant shifts in consumer preferences. To achieve this, mathematical modeling is used to design a distribution system that maximizes profits in a competitive market, comprising a wholesaler and multiple retailers across multiple periods and products. Customer behavior is captured through a behavior-based pricing process at the retail level, with equilibrium values determined using bi-level programming based on Stackelberg modeling, which accounts for asymmetric competition. The model is solved using two distinct approaches: structural modification and data-driven learning models. In the structural modification approach, the bi-level model is linearized and converted into a single-level equivalent. Meanwhile, in the data-driven approach, the pricing process is managed using the CLIQUE clustering method, which helps develop a rule-based pricing system grounded in data extraction. Numerical examples and sensitivity analyses are provided to illustrate the concepts, and the outcomes are compared to highlight managerial implications and avenues for future research.

Developing and optimizing effective inventory systems considering realistic constraints and practical assumptions can help managers remarkably decrease inventory and consequently supply chain costs. In this research, we propose a new variant of the multi-item inventory model taking into account warehouse capacity, on-hand budget constraints, imperfect products in supply deliveries and partial backordering where the products can be converted into perfect products by a local repair shop. To deal with the proposed model, three solution approaches, including interior-point technique, as an exact method, and two metaheuristics based on Simulated Annealing (SA) and Water Cycle Algorithm (WCA), are proposed. Extensive computational experiments are conducted on different sets of instances. Using different measures such as RPD, PRE, and computational time, the performance of the solution approaches is evaluated within different test instances. The results show that the WCA outperforms the two other approaches and leads to the best solutions in the proposed problem.

Efficient flow management is critical in transportation and logistics, with hub networks playing a key role in optimizing these processes. The hub arc location problem has recently emerged as a new framework that emphasizes hub arcs while allowing for isolated hubs. This paper extends the hub arc location problem by incorporating set-up costs into the optimization model. A heuristic algorithm is developed to enhance hub network design, considering both the flow of goods and the associated hub set-up costs. Additionally, a detailed sensitivity analysis is conducted to assess the impact of strategic adjustments on optimization outcomes. By reducing the discount factor for inter-hub flows and increasing the number of exogenous hub arcs, significant improvements in route optimization and cost reduction are achieved. This research challenges traditional approaches to hub network design and opens the door for further exploration of the dynamics within hub networks. A deeper understanding of these networks can lead to more efficient and resource-optimized transportation systems, potentially transforming flow management into a more cost-effective and sustainable process.

This study utilizes deep-learning models for stock price prediction, focusing on data from five companies listed on the Tehran Stock Exchange over the period 2001 to 2022. Five models are employed, including two hybrid models and three single models. The hybrid CNN-LSTM model serves as the primary model, with its predictive accuracy compared against the other four models. Results indicate that the CNN-LSTM model demonstrates superior performance relative to the others, although the CNN-GRU hybrid model also yields satisfactory results. Interestingly, among the single models, the CNN model surpasses both the LSTM and GRU models, defying initial expectations. The accuracy of the models is notably impacted by factors such as volatility, which increases uncertainty. This research, which exclusively relies on technical indicators, suggests that achieving optimal results hinges not only on selecting the right neural network but also on determining the appropriate number of layers in each model. Overall, the CNN-LSTM model delivers the best performance across four of the five stocks, with the CNN-GRU model slightly outperforming it for one stock. Among the single models, the CNN model consistently outperforms the others.