Abstract: This work presents an advanced particle swarm optimization algorithm featuring two decoding schemes, namely PSODS-I and PSODS-II, for solving the Vehicle Routing Problem with Time Windows (VRPTW). Both schemes begin by generating a customer precedence list, followed by the formation of a vehicle precedence list and subsequent route construction. To further enhance solution quality and reduce transportation cost, an intelligent local search mechanism is developed based on a heuristic improvement strategy. The proposed schemes are evaluated on Solomon’s benchmark instances with 25 and100 customers using key performance indicators, including total travelled distance and number of vehicles. In addition, the consistency and robustness of the solutions are assessed and compared with the best known solutions and several state-of-the-art methodologies. Comparative analysis demonstrates that the proposed approach, particularly PSODSII, is highly competitive in terms of both fleet size and total distance travelled for solving the VRPTW.

Abstract: There are many uncertainties in the process of technology projects “Unveiling and Commanding” systems which can easily lead to the risk of failing to achieve the expected effect of technology quality. In order to ensure that the implementation of the system can achieve the expected results, the indicator system for risk assessment of technology projects “Unveiling and Commanding” system is established on the basis of expert interviews and field research. The improved game theory multiple combination weighting method is used to weight the indicators to reduce error risk of the weight results. This paper makes a comprehensive risk assessment of technology projects “Unveiling and Commanding” systems from the two dimensions of risk possibility and risk impact degree based on two-dimensional cloud model, so as to reduce the subjectivity, randomness and fuzziness of assessment results. The example verifies that the model is reasonable and effective, and has a certain guiding role in the improvement of technology projects “Unveiling and Commanding” system.

Abstract: With the deepening of power market reform, small and medium-sized users have gradually been included in the agency power purchase mechanism. The contradiction between price fluctuation risks, load uncertainty and cost control faced by these users has become increasingly prominent. These users are numerous, but their loads are scattered and their risk tolerance is weak. The traditional static electricity price model is difficult to adapt to the dynamic changes of the market and the differentiated demands of users. Therefore, this paper proposes a dynamic adjustment framework for the agency power purchase electricity prices of small and medium-sized users, which includes market perception, risk quantification, optimization decision-making, and execution feedback. This framework is based on the price risk quantification model, multi-objective optimization function, and improved dynamic weight algorithm, integrating historical price trend weighting, risk cost trade-off, and real-time load response feedback, to achieve the collaborative optimization of power purchase costs, risk losses, and user satisfaction. Experimental results show that this method performs well in three typical scenarios of stability, fluctuation, and extreme conditions: the price risk loss is reduced by 32.7% to 41.2%, the average power purchase cost is stable at 0.38 yuan/kWh to 0.42 yuan/kWh, user satisfaction reaches 92.3%, and the comprehensive performance is significantly superior to the traditional fixed electricity price method, single cost optimization method, and static risk control method. In scenarios of price sudden change and large load peak-valley difference, the risk loss is still lower than 8.5%, the response delay is controlled within 50 ms, demonstrating good robustness and real-time performance. This research provides an efficient and adaptive electricity price adjustment technical path for agency power purchase of small and medium-sized users, which can be applied to various types of power market environments, differentiated user groups, and dynamic load demands.

Abstract: Large earthquakes often disrupt road networks, severely hindering the timely delivery of emergency supplies. This paper studies a truck–UAV collaborative emergency routing problem under road blockage conditions. We develop a mixed-integer programming model that coordinates truck and UAV operations to minimize total emergency response time while accounting for payload, endurance, demand satisfaction, and road repair constraints. To solve the problem efficiently, we propose an improved Variable Neighborhood Search algorithm with greedy initialization (VNS-G), together with a benchmark variant based on random initialization (VNS-R). Computational experiments on instances of different sizes are conducted to evaluate the proposed approach. The results show that VNS-G can obtain high-quality solutions close to those of CPLEX on small-scale instances. On medium-scale instances, it provides substantial computational savings while maintaining competitive solution quality, reducing computation time from 682–3496 s for CPLEX to 95–340 s in the tested cases. For large-scale instances, where exact optimization becomes computationally impractical, the proposed heuristic remains effective in generating feasible solutions within operationally meaningful time. Compared with VNS-R, VNS-G generally achieves better solution quality and search performance. A case study of the earthquake-prone Ya’an region further illustrates the model's practical applicability. Sensitivity analysis reveals a mechanism-based managerial insight: UAV endurance primarily affects reachability, whereas UAV payload more directly improves response efficiency by reducing the need for repeated sorties. This suggests that, once reachability is ensured, improving payload capacity is likely to generate greater operational benefits than further extending endurance.

Abstract: Optical film cutting is a critical upstream process in LCD and electronics manufacturing, characterized by high material costs, diverse order specifications, and frequent demand fluctuations that complicate production planning. In practice, manufacturers must coordinate cutting methods, manage semi-finished inventory across multiple periods, and handle recyclable waste simultaneously. However, these decisions are often made independently, resulting in excessive material usage, unstable inventory levels, and avoidable waste. This study develops a multi-period optimization framework for LCD optical film cutting that integrates two production routes, namely slitting and miter cutting, with inventory control and waste recycling. The model captures key operational trade-offs, including process selection, timing of intermediate production, and recycling decisions. It first minimizes total costs, incorporating material, processing, inventory, and waste-related costs, and is further extended into a bi-objective formulation that jointly considers operational cost and waste generation. The primary contribution of this research is to provide an integrated decision framework that reflects real production conditions in electronics manufacturing. By linking cutting, inventory, and recycling decisions across multiple periods, the model enables more coordinated planning and improved resource utilization. A case study with large-scale order data shows that the proposed approach can reduce waste, stabilize inventory, and improve the balance between cost efficiency and environmental performance. The analysis also reveals that cutting assignments tend to concentrate around several preferred angle configurations under realistic production conditions, suggesting the existence of practical process preferences and recurring operational patterns in optical film manufacturing. Overall, this study offers a practical planning tool for optical film converting operations and is applicable to other high-value, roll-based materials in precision manufacturing environments where inventory linkage, process flexibility, and recyclable waste are critical considerations.

Abstract: Centralized alliance procurement exerts a profound impact on pharmaceutical enterprises innovation, which is an essential pathway for pharmaceuticals to enter hospitals in China. In order to explore whether alliance procurement inhibits or promotes innovation transformation of pharmaceutical enterprises, this study constructs an evolutionary game model between local government and pharmaceutical enterprises. Based on the data of some provinces’ centralized procurement platforms, online pharmacies and China Drug Administration, the model parameters were assigned, and the dynamic evolution and influencing factors of pharmaceutical enterprises’ innovation behavior under alliance procurement were studied through simulation. The results show that the buyer power generated by alliance procurement affects the innovation decision of pharmaceutical enterprises, and the increase of buyer power promotes the innovation of pharmaceutical enterprises to a certain extent. When buyer power exceeds the threshold value, pharmaceutical enterprises tend to innovate in the short term, but stabilize at a non-innovation strategy in the long term. At the same time, research and development cost, initial willingness to innovate, innovation subsidies and other factors affect the speed of enterprises to stabilize the strategy. This study provides insights for the government to promote the innovation transformation of pharmaceutical enterprises under centralized procurement.

Abstract: AI technology has become a key means to improve product quality information disclosure, alleviating information asymmetry in supply chains. This study investigates a two-echelon supply chain in which a single supplier interacts with two competing retailers that are differentiated in terms of product quality. Four AI adoption strategies are considered, including non-adoption of AI by both retailers (NN), AI adoption only by the high-quality retailer (AN), AI adoption only by the low-quality retailer (NA), and AI adoption by both retailers (AA). We further analyze how key factors, such as quality competition intensity, AI efficiency coefficient, AI investment level and information asymmetry degree, affect supply chain decisions and optimal AI adoption strategies of all supply chain members. The results reveal that the high-quality retailer always benefits from AI adoption and gains the highest price premium under the AA strategy. In contrast, the low-quality retailer can achieve positive profits only when the fixed cost of AI platforms is below a critical threshold. The supplier achieves optimal profits under the AA strategy with moderate competition and under the NA strategy with intense competition. Core parameters such as the AI efficiency coefficient and information asymmetry jointly influence corporate AI adoption decisions, and the NN strategy is preferred amid severe information asymmetry.

Abstract: This study presents a Reinforcement Learning (RL)-assisted Genetic Algorithm (GA) framework for the Multi-Objective Capacitated Vehicle Routing Problem with Time Windows (MOCVRPTW). In this problem, a set of homogeneous vehicles depart from a depot, visit all customers exactly once, and return back to the depot. The routes of the vehicles are constructed by considering three objectives: minimizing the total travel time, minimizing the number of vehicles, and maximizing the satisfaction obtained from the customers who are visited within their time windows. We propose using an NSGA-II-based approach that is assisted by Q-learning-based operator selection methods for this problem. Unlike traditional GAs that use fixed operators, the proposed approach enables learning-based selection of each operator (crossover and mutation) considering the current performance of solutions. We make tests with five different Q-learning-based operator selection strategies and compare their results to using fixed or randomly selected operators by nonparametric statistical methods. The results show that all Q-learning-based operator selection strategies outperform the fixed-operator approach, whereas the random selection strategy is outperformed by four. In addition, when the best operator for each state of solutions is found considering all solution approaches and used in NSGA-II throughout the algorithm, it consistently results in the best performance among all. Overall, the results demonstrate that the proposed RL-supported GA framework provides a competitive alternative in terms of Pareto-front solution quality for MOCVRPTW, and learning-based operator selection can be an effective mechanism for adaptively controlling the evolutionary search process.

Abstract: Public-private cooperation is vital for emergency stocks, yet complex behavioral characteristics of governments and suppliers often affect system reliability. Existing research has mainly examined supplier loss aversion, with limited attention to government inequity aversion. This paper investigates how to strategically reserve and procure emergency supplies under the roles of supplier loss aversion and government inequity aversion. We develop an option contract-based emergency stockpiling model that incorporates a loss-averse supplier and an inequity-averse government. Moreover, comparisons between the decisions yield the boundary conditions for supplier understocking and overstocking. The conditions for humanitarian channel coordination under both disadvantageous and advantageous inequity aversion are derived, with some overlapping and others differing between both situations. A case study using laboratory simulations is conducted. The results indicate that higher loss aversion and optimism levels decrease the supplier’s best reserve quantity; disadvantageous (advantageous) inequity aversion increases (reduces) the government’s optimal purchase of physical options. The boundary conditions for supplier understocking and overstocking depend on the interaction between government inequity aversion and supplier loss aversion. Finally, some targeted managerial insights are proposed to coordinate the humanitarian supply chains, select appropriate suppliers, and evaluate the behavioral preferences of participating members.

Abstract: The continuous advancement of artificial intelligence technology is impacting the operational mechanisms and market competition structures of closed-loop supply chains. How to integrate AI empowerment with channel strategies for coordinated optimization has become a core decision-making issue for manufacturers. By establishing three progressive game models: the benchmark model (M0), the encroachment model (M1), and the AI empowerment model (M2), this study investigates the strategic interplay arising from manufacturers' channel encroachment decisions and AI technology deployment in the context of remanufacturing outsourcing, along with the resulting economic implications. The study finds that when channel differentiation is significant, the manufacturer's encroachment can enhance its profits through market expansion effects. On this basis, by introducing AI technology, its enabling effect exhibits a non-linear characteristic. Only when the channel substitution degree is low and the price sensitivity is moderate, can AI achieve the synergistic benefits of cost savings and market expansion. In terms of revenue distribution, the AI dividend has extensive penetration among supply chain members. Retailers only suffer losses in extreme scenarios where the channels are highly homogeneous and consumers are not price-sensitive. In most market conditions, they can share the efficiency improvement brought about by technology empowerment. This study theoretically reveals the interaction mechanism between channel encroachment and AI empowerment, expands the analytical boundaries of closed-loop supply chain management, and provides strategic guidance for manufacturing enterprises on how to formulate collaborative strategies in the wave of digital transformation.

Abstract: With the rapid development of electric vehicles, power battery recycling and utilization have attracted increasing attention. This study employs game theory to investigate power battery recycling subsidy strategies under two distinct recycling models (i.e., retailer-only recycling and joint recycling by the retailer and the third-party recycler), as well as the impact of recycling models on the subsidy strategies. The results reveal that when the required recycling subsidy level is high, the government chooses not to subsidize; when the required recycling subsidy level is low, the government provides subsidies, but subsidy strategies vary significantly under different recycling models. Specifically, when the manufacturer recovers power batteries only through the retailer recycling channel, subsidizing any supply chain member can maximize social welfare. By contrast, when the manufacturer adds a third-party recycling channel, the government should provide subsidies to the manufacturer or consumers to maximize social welfare. Furthermore, the retailer prefers government subsidies for itself in the retailer–third-party joint recycling model, but not in the retailer-only recycling model. In addition, this study examines the impacts of key parameters on the firms' profits and social welfare, and shows that as the government subsidy increases, the profits of supply chain firms rise, while social welfare first increases and then decreases. These findings can support decision-making by firms and help governments formulate relevant policies for closed-loop supply chains.

Abstract: In the last few years, artificial intelligence has grown rapidly and has become increasingly popular in daily life. As a powerful branch of artificial intelligence, generative AI (Gen AI) offers immense potential for application within the new product development (NPD) process. In this study, we investigate the impact of Gen AI on NPD by utilizing a closed-form analytical model within a single-product monopoly. The results show that the profit and productivity levels of a company adopting Gen AI technology are lower than those of a non-Gen AI company. However, the company that adopts Gen AI technology achieves better product quality than a company without Gen AI implementation. Furthermore, by increasing design costs by a small amount, the Gen AI adopting company improves its product quality level, while non-Gen AI companies experience only a slim increase in product quality. Additionally, under lower fixed costs, outcomes improve for both Gen AI and non-Gen AI companies. Our findings suggest that a company which has already adopted Gen AI technology should not reduce the cost of quality. Instead, it should focus more on increasing the cost of quality to realize greater benefits from Gen AI implementation.

Abstract: In the case of disruptions in the blood supply chain, rapid reorganization of distribution processes is required to ensure an effective response to emergency situations. This paper considers the problem of redistributing available blood stocks from the institute and hospitals to hospitals affected by a disruption. A mathematical formulation of the problem is developed, with a multi-objective function aiming to: (i) minimize the blood delivery time to the locations of disruption, (ii) minimize violations of predefined safety stock levels at the institute and hospitals, and (iii) minimize the amount of blood taken from hospitals not affected by the disruption. The formulation also introduces constraints that ensure balanced violations of safety stock levels across unaffected facilities. Computational experiments are conducted on test scenarios generated from real case studies from the healthcare system of the Republic of Serbia. Small-sized instances can be solved exactly, providing benchmarks for evaluating a General Variable Neighborhood Search metaheuristic designed for larger problem instances. The results indicate that the proposed metaheuristic produces high-quality solutions within negligible CPU time.

Abstract: Digital transformation has become a pivotal driver of omni-channel retailing advancement, while simultaneously reshaping consumer purchasing behavior. In response to this shift, an increasing number of retailers have adopted Buy-Online and Pick-Up-in-Store (BOPS) as a strategic instrument for omni-channel integration, aiming to enhance shopping convenience and strengthen profitability. Moreover, given the inherent uncertainties of online shopping, retailers increasingly leverage online reviews to provide consumers with more product information, thereby facilitating informed purchase decisions. Two analytical models are formulated to characterize the scenarios before and after BOPS implementation. A comparative analysis is subsequently conducted to evaluate the effects of BOPS on the retailer’s pricing, demand, and profitability. Furthermore, this study examines the influence of online reviews on the retailer’s operational performance and investigates their interactive effects with product returns on retailer profitability under both models. The results demonstrate that the introduction of the BOPS channel enhances retailer profitability when at least one of the unit operating cost or return processing cost is low, while the other is low or moderate. Furthermore, the effectiveness level and the weight of online reviews are critical determinants in the retailer’s decision-making process. The retailer should retain online reviews when the effectiveness level is high, or when it is low but the weight is high. Finally, considering the combined impact of online reviews and product returns, the analysis suggests that the retailer can achieve enhanced profitability through the adoption of BOPS under specific conditions.

Abstract: To address the issues of premature convergence and insufficient optimization accuracy in existing metaheuristic algorithms, this study proposes a novel Strawberry Runner Propagation Algorithm (SSPA). The algorithm is inspired by the biological process of strawberry stolon growth and establishes a four-stage optimization framework consisting of exploration, transition, exploitation, and convergence. In the exploration stage, Levy flight and Sobol sequence–based sampling enhance population diversity and global search capacity. The transition stage employs differential evolution and gradient approximation strategies to enable adaptive and intelligent state adjustments. The exploitation and convergence stages utilize centripetal convergence and multi-elite weighted fine-tuning to strengthen local optimization and improve final solution precision. The performance of SSPA was evaluated on 29 benchmark functions in the CEC2017 test suite and compared with representative metaheuristic algorithms. Experimental results show that SSPA achieves faster convergence, higher optimization accuracy, and stronger robustness. Furthermore, its application to robotic path planning in complex obstacle environments demonstrates significant improvements in path quality and computational efficiency. These results indicate that SSPA provides an effective and generalizable optimization framework for complex engineering and intelligent planning tasks.

Abstract: By building a game model between brands and outsourced anchors, this article systematically studies the incentive design, opportunistic behavior and blockchain governance mechanism in live streaming. We found that while promoting the sales efforts of anchors, commission incentives will also induce opportunistic behaviors such as AI rhetoric exaggeration, traffic manipulation and channel encroachment. Secondly, blockchain technology can effectively suppress opportunism by improving the traceability of information and the verifiability of behavior: on the one hand, it increases the marginal cost of implicit violations, On the other hand, contractible punishment is imposed on explicit sexual encroachment. The key threshold conditions in the study show that when the sum of the incentive intensity and the blockchain punishment exceeds the encroachment of personal interests, the encroachment can be completely suppressed. The study further reveals the three-state equilibrium model of the optimal blockchain governance intensity: strong governance, weak governance and precise governance. Its choice depends on the trade-off of product characteristics, incentive costs and governance benefits. At the same time, the governance effect is affected by the power structure of the platform, and the platform, as a provider of governance infrastructure, may have the problem of insufficient motivation. Based on this, this article proposes that the sustainable development of live ecology requires building a comprehensive system of incentive contracts, technical governance and institutional environment coordination. It also provides a theoretical framework for understanding the problem of entrusted agencies in digital marketing and a decision-making basis for the governance practice of the platform economy.

Abstract: This study investigates the mixed-model two-sided U-type assembly line balancing (MMTsUtALB) problem and proposes a hybrid solution approach based on Genetic Algorithm (GA) and Variable Neighborhood Search (VNS). Unlike existing studies on two-sided U-type assembly line balancing (TsUtALB), the mixed-model structure is explicitly considered. In the proposed approach, GA is used to explore the solution space through evolutionary operators, while VNS is applied as a local improvement procedure to refine promising solutions and improve convergence. A problem-oriented priority rule–based encoding method is adopted to represent solutions, which are transformed into feasible two-sided U-type assembly line configurations using a decoding-based task assignment procedure. This structure allows the algorithm to balance diversification and intensification during the search process. The effectiveness of the GA–VNS hybrid algorithm is evaluated using benchmark test problems. Since this study represents the first attempt to solve the MMTsUtALB problem, the obtained results are compared with closely related mixed-model two-sided assembly line balancing studies. Computational results show that GA–VNS achieves competitive performance, particularly for larger instances, by reducing the number of stations and positions with lower computational times than SA and PSO.

Abstract: Industrial Control Systems (ICS) are critical infrastructure components that manage essential services including power grids, water treatment facilities, and manufacturing processes. The Modbus TCP protocol, widely deployed in these systems, lacks inherent authentication mechanisms, rendering it vulnerable to replay, man-in-the-middle, and impersonation attacks. Existing authentication solutions based on public-key cryptography impose significant computational overhead on resource-constrained Programmable Logic Controllers (PLCs). In this paper, we propose a lightweight bidirectional authentication scheme for Modbus TCP utilizing enhanced Chebyshev chaotic maps with binary-exponentiation-based acceleration. We formally analyze the protocol in the Dolev-Yao model using ProVerif and evaluate its computational and communication costs on representative industrial computing platforms. The results show a total authentication cost of approximately 3.11 ms, including 1.17 ms on the slave side, with a communication overhead of 1280 bits across four messages. Based on our literature review, the contribution of this work is not the use of chaotic maps per se, but their adaptation to the Modbus TCP setting through a Modbus-oriented credential structure, low slave-side computational burden, and a deployment path compatible with existing protocol stacks. Direct validation on commercial PLC hardware remains an important next step.

Abstract: Digital Twin Shop-floor (DTS) systems require effective representation and integration of heterogeneous manufacturing data to support real-time decision-making. Existing approaches often suffer from limited semantic connectivity and insufficient support for dynamic data updates, restricting their applicability in complex manufacturing environments. To address these challenges, this paper proposes a DF-MKG-DTS framework that integrates a Manufacturing Knowledge Graph (MKG) into a Data Fabric (DF) architecture. The framework enables unified representation, semantic association, and dynamic updating of multi-source shop-floor data by modeling manufacturing entities, relationships, and temporal attributes as a structured Knowledge Graph, thereby improving fault localization accuracy. Experimental results demonstrate that the data association rate reaches 99.5% before knowledge fusion and achieves full association after fusion, validating the effectiveness of the proposed approach. For small-scale temperature faults, the proposed framework achieves 92.6% accuracy, an improvement of 10.2% over manual inspection, and maintained 89.8% accuracy for large-scale, complex process combination faults while reducing average processing time by up to 40%. These results indicate that DF-MKG-DTS not only alleviates data isolation and enhances data utilization but also provides robust, scalable, and efficient fault diagnosis support, highlighting its practical value for Digital Twin Shop-floor applications.

Abstract: Multimodal transport hubs (MTH) improve the operations between different transport modes, creating quasi-seamless connections between origin-destination locations. These hubs can fully use the advantages of various modes of transportation and avoid the limitations and cost pressures caused by a single mode. However, locating where to situate these hubs is a complex and critical process. Recent site location selection studies only assume that the hub is unstable under a single transportation mode, with uncertain demand and uncertain costs, which is very different from reality. This work approaches the hub selection problem following a particle swarm-simulated annealing algorithm constrained by transportation demand and cost. A case study was conducted in Northeast China, selecting six hubs responsible for about 93% of the cargo flow in the network. Our results suggest that selecting transportation demand and cost for a comprehensive analysis of hub location selection will optimize the hub layout and reduce the total system cost. The scale of freight efficiency would be improved compared to the existing transportation network.

Abstract: To alleviate the problem of long waiting time in general hospitals and underutilization of community hospital resources, this paper explores the effect of government mixed subsidies on patient choice and utilization of community hospital resources. Based on hospital service quality, we establish a benchmark model without government subsidy and a bilevel optimization model in which the government makes the upper-level decisions while the community hospital responds at the lower level. Although the benchmark model is a nonconvex optimization model, we are based on its special structure to propose a globally optimal algorithm. Then, we combine this globally optimal algorithm with the Gauss-Seidel method to solve the bilevel optimization model. Furthermore, we present a sensitivity analysis of some key parameters such as service quality costs, basic medical expenses in community hospitals, and government subsidy budget. Theoretical and numerical results provide the following insights: Government mixed subsidy may improve community hospital service quality and so attract more patients to choose community hospital; Government should focus on fixed fund subsidies to improve community hospital service quality, since hospital service quality is more critical than medical expenses for patient choices; Controlling service quality costs and basic medical expenses in community hospital at a low level and paying more attention to fund subsidies can promote hierarchical healthcare. Meanwhile, the government subsidy budget should be decided aligning with defined societal goals.

Abstract: Carbon taxation is a pivotal policy tool for steering manufacturing toward circular economy models, yet its effectiveness hinges on how market forces respond. This study reveals a critical but overlooked link: the strategic marketing efforts of downstream retailers. We develop a Stackelberg game model of a closed-loop supply chain where a manufacturer produces both new and remanufactured products under a carbon tax, and a retailer invests in marketing to promote remanufactured goods. Our analysis demonstrates that while carbon taxes inherently boost the competitiveness of low-carbon remanufactured products, this “green signal” is significantly amplified by retailer marketing. However, decentralized decision-making leads to a dual inefficiency: classic double marginalization and chronic underinvestment in marketing due to its positive externality, thereby dampening overall policy effectiveness and supply chain performance. To resolve this, we design and validate a novel hybrid “Cost-Sharing and Revenue-Sharing” (CS-RS) contract. This mechanism perfectly coordinates the chain by internalizing the marketing externality, aligning incentives, and ultimately restoring the profit and environmental performance of a centralized system. Numerical simulations further show that stronger consumer green preferences enhance both the impact of carbon taxes and the value of coordination. Our findings contribute a unified “policy-production-marketing” framework to sustainable operations literature. For managers, we provide a clear pathway to transform carbon compliance into competitive advantage through upstream-downstream collaboration. For policymakers, we highlight the necessity of considering supply chain dynamics and market-building initiatives to ensure environmental regulations achieve their intended impact.

Abstract: As platform economies become increasingly intelligent, platform provided artificial intelligence (AI) services have emerged as an important tool for brand manufacturers to improve competitiveness. This study examines whether platform AI serves as a blessing or a curse for brand manufacturers. We develop a benchmark model with a single manufacturer and extend it to two Stackelberg game settings featuring either similar market positions or a strong and a weak manufacturer. We analyze how strategic AI adoption shapes pricing decisions, firm profits, consumer utility and social welfare. The results show that without a competition platform AI increases the manufacturer’s profits, thus constitutes a clear blessing for the brand manufacturer. When competition exists brand manufacturers always have incentives to adopt platform AI and unilateral adoption benefits the adopter. However, under Nash equilibrium where both manufacturers adopt AI outcomes differ by market structure. With symmetric competition equilibrium adoption leads to a prisoner’s dilemma that harms both manufacturers. With asymmetric competition the prisoner's dilemma affects only the weak manufacturer while the strong manufacturer continues to benefit making platform AI a curse for the weak party. We further show that under symmetric competition equilibrium AI adoption always improves platform profits, consumer utility and social welfare. Under asymmetric competition consumer utility still rises but the effects on platform profits and social welfare depend on manufacturers costs and the platform commission rate. Extensions indicate that asymmetry in base demand shares does not alter equilibrium adoption. Overall, the study offers managerial insights into AI adoption in competitive platform markets.

Abstract: Focusing on a contract farming supply chain empowered by the low-altitude economy, this paper constructs a Stackelberg game model composed of an acquisition firm and a farmer. By incorporating key variables such as low-altitude investment costs, technology transformation efficiency, and consumer technology preference, we comparatively analyze the optimal decisions and social welfare under three scenarios: no government subsidy, government subsidy to the farmer, and government subsidy to the firm. The results indicate that: First, regardless of whether the government subsidy recipient is the farmer or the firm, the optimal output, low-altitude investment level, and profit distribution among supply chain members are identical in equilibrium; the policy difference lies solely in the timing and cost of fund execution. Second, a reduction in low-altitude investment costs or an improvement in technology transformation efficiency is not always beneficial to the farmer. An excessively high technology substitution rate grants the firm strong bargaining power, enabling the firm to capture technology dividends by suppressing the procurement price, thereby causing the farmer's profit to exhibit an inverted U-shaped trend of increasing first and then decreasing. Third, government subsidies are not always effective. When consumer preference for new technology agricultural products is high or technology costs are moderate, the market mechanism can already achieve optimal allocation; consequently, intervention at this stage would suppress total social welfare. The conclusions of this study provide a theoretical basis for the government to optimize agricultural subsidy strategies regarding the low-altitude economy and for the firm to formulate sustainable procurement mechanisms.

Abstract: With the continuous enrichment of live-streaming content, consumers gradually exhibit different preferences for live streaming shopping. This study constructs a centralized and decentralized supply chain game model between a supplier and a live-streaming enterprise (LSE), considering shopping-preferred consumers and content-preferred consumers. We find that in centralized and decentralized settings, the total profit for the content-preferred consumers is always higher (lower) than for the shopping-preferred consumers when the selling conversion rate of traffic is low (high). Furthermore, a lower traffic cost coefficient does not always result in a lower product price in a centralized setting compared to a decentralized setting. Only a higher traffic cost coefficient will lead to a lower price in a centralized structure compared to a decentralized setting. Finally, we design a two-part tariff contract and a wholesale price rebate contract based on traffic with fixed transfer payments. Our findings fill a gap in the supply chain coordination literature. Moreover, we provide marketing strategies for suppliers on selecting LSEs and for LSEs on positioning consumer preferences, while offering firms a to coordinate contract options.