Target Service Level Research Articles

EXPRESS: Expedited Shipping to Meet a Target Service Level: Analytical Recommendations and Behavioral Biases

EXPRESS: Expedited Shipping to Meet a Target Service Level: Analytical Recommendations and Behavioral Biases

Inventory Management Improvement at PT Inalum, Kuala Tanjung, Sumatra Utara

This research aimed to analyze the inventory issue in PT Inalum. The method used in this research is a quantitative method to collect inventory historical data from January 2022 to December 2022 obtained from the company’s ERP SAP. The data contains information of 6028 items of the MRO inventory. The collected data was analyzed using Economic Order Quantity (EOQ), Reorder Point (ROP), Safety Stock (SS) and Average Inventory Level (AIL). By using Fixed Order Quantity model, the MRO inventory is calculated to compare the ideal and actual average inventory levels. From the analysis, the researcher found that the MRO inventory in Kuala Tanjung, PT Inalum, has a service level of 0.52, which is still far below the service level target of 0.95. At the same time, PT Inalum has the potential to reduce inventory by USD 5,402,389 or 58.2% of the inventory value. This research aims to provide guidance for the management of PT Inalum in making inventory policies that can increase service levels as well as minimize inventory costs.

A multi-level production-inventory-distribution system under mixed make to stock, make to order, and vendor managed inventory strategies: An application in the pulp and paper industry

In today's competitive environment, most companies tend to adopt a hybrid manufacturing strategy to provide more reliable services to a wide range of customers. Based on a real-world case study, we investigate a customer-wise production-inventory-distribution system in a multi-level supply chain. Within this system, at a tactical planning level, the decision concerning the production strategies including make to stock (MTS), make to order (MTO), or vendor managed inventory (VMI) strategy, to choose for each customer is made, and then at the operational planning level, the production, inventory, and distribution activities are coordinated according to the strategy assigned to the customers. The objective is to optimize the total cost while the VMI customers are timely satisfied and the target service level for MTO customers is respected. Furthermore, the purchase acceptance rate for MTS customers can be controlled to enhance capacity utilization. We develop a rolling horizon planning approach to deal with the dynamics of customers' demands under different forecasting scenarios. We perform numerical experiments under both static and rolling horizons based on adapted data from a real-world application from a Canadian pulp and paper industry. The results reveal that customers placing smaller orders are classified within the VMI strategy to mitigate potential delays. Conversely, MTS customers are predominantly characterized by larger order quantities. A sensitivity analysis of key problem parameters demonstrates that the backorder cost for MTO customers can effectively influence the strategy allocation. Moreover, our results underscore the notable influence of sales prices on the distribution of customers among strategies.

Retail store-SKU level replenishment planning with attribute-space graph recurrent neural networks

Large retail chains routinely make replenishment decisions for inventory at the store Stock-Keeping-Unit (SKU) level, relying heavily on accurate demand forecasting to meet target service levels. We critically examine the current state-of-the-art in demand forecasting, revealing a notable gap in existing models, specifically the neglect of intricate cross-SKU promotional effects. To address this gap, we introduce an innovative solution: the Attributed Space Graph Recurrent Neural Network (ASG-RNN), which encodes demand history and drivers' history using a recurrent neural network, and captures cross-SKU promotional interactions through a graph convolution network. The graph convolutions operate on an attribute space graph constructed using K-nearest neighbor algorithm based on SKU-level static attributes. Initial empirical tests were conducted on a dataset comprising 200 SKUs in the Coffee category from a retail chain with 93 stores. The analysis demonstrates that ASG-RNN, on average, reduces loss by 4.1% compared to the best benchmark, resulting in average inventory cost savings of 4.94%. Notably, for promoted SKUs, ASG-RNN reduces loss by 13% on average, and it can achieve a 30% reduction in loss when the target service rate is high. Further experiments conducted on four additional product categories, including Beer, Laundry Detergent, Milk, and Mayo, yield similar results, showcasing the robustness of ASG-RNN's performance. These findings underscore the practical significance and cost-effectiveness of our innovative approach in enhancing demand forecasting and optimizing replenishment decisions for large retail chains.

Distributionally robust chance-constrained optimization for the integrated berth allocation and quay crane assignment problem

In this paper, we are the first to propose a distributionally robust chance-constrained (DRCC) optimization model for the integrated berth allocation and quay crane assignment problem (BACAP) in container terminals. In contrast to the classical deterministic BACAP model, we consider the arrival time to be uncertain due to the frequent arrival delays in ports. We then impose a chance constraint that the service time must start after the uncertain arrival time with a probability of at least 1−ϵ, where 1−ϵ represents the target service level in container terminals (ϵ represents the target risk tolerance of the port manager). Under the moment-based ambiguity set, we reformulate the DRCC model into a mixed integer semi-definite programming (MISDP) model. Additionally, we develop an efficient decomposition branch-and-bound algorithm to solve the MISDP model and obtain the exact solution. Fortunately, a special case of the DRCC model arises when the mean and covariance utilized in the ambiguity set are precise, allowing for the transformation of the DRCC model into a mixed integer programming (MIP) model. This conversion significantly reduces the complexity of the problem. Impressively, the solving time of the MISDP model with the decomposition branch-and-bound algorithm is comparable to that of the transformed MIP model. The numerical results show that our model can achieve a schedule with high service at low cost. Meanwhile, we have made an intriguing discovery that the correlation between the target risk tolerance and the actual service level can be depicted as a staircase function regardless of the datasets. This finding offers crucial insights for port management, enabling them to strike a balance between the cost and actual service level by determining an appropriate target service level.

Fourth-party logistics network design with demand surge: A greedy scenario-reduction and scenario-price based decomposition algorithm

The sharing economy has offered new viable approaches for firms to tackle demand surges. In this paper, we study a novel network design problem for the fourth-party logistics network (4PLN) to cope with unpredictable demand surges, in which temporary renting of logistics resources from a sharing market is available to supplement network capacity. A chance-constrained stochastic programming model is established to minimize the overall cost for 4PLN under service level targets. To deal with the difficulties brought by the evaluation of service level, we reformulate a mixed-integer linear programming (MILP) model, that can be solved straightly, based on the sample average approximation method. Then, to address the enormous challenge posed by the coupling of the basic NP-hard network design problem and the large number of demand scenarios in the MILP version, the Scenario-Price based Decomposition Algorithm (P-DA) is designed based on the key idea of decomposing the above-coupled factors. Further, to mitigate the performance deterioration brought on by large system scale and/or sample size, we expand our base algorithm to the Greedy Scenario-Reduction and Scenario-Price based Decomposition Algorithm (GR&P-DA) through the fast processing of chance constraints by introducing a greedy method. Through computational experiments, we demonstrate the effectiveness of our proposed model and algorithms, and we analyze the impact of various model parameters, including demand level, surge magnitude, surge probability, and rental price of third-party logistics resources, on 4PLN design. Furthermore, our comparative analysis reveals that renting logistics resources to address demand surges can significantly reduce costs and discovers that deploying resources in advance does not always provide an advantage, a temporary “case-by-case” planning approach for renting will give a more cost-saving scheme when surge probability is at a low level as it avoids idle resources by not being stuck with a conservative strategy.

Supply chain modelling and LARG performance evaluation under severe disruptions: the case of the fast-food industry

the LARG paradigm involves and combines different perspectives: Lean, Agile, Green, and Resilience. These four perspectives are recognized as a suitable response to the strong demand fluctuations and characteristics of modern variables and unpredictable markets. Their integration into modern supply chains is also crucial from a strategic point of view. This study develops a simulation model useful for measuring and evaluating the LARG performance of a supply chain operating in the fast-food industry. Different configurations have been tested varying the number of players and the target service level. Then, different scenarios with distortions have been compared. Results show that the service level influences the LARG performance regardless of the kind of disruption. The supply chain structure can affect LARG performance when the demand increases, while it seems not relevant if the demand decreases.

Empowering the Capillary of the Urban Daily Commute: Battery Deployment Analysis for the Locker-Based E-bike Battery Swapping

In densely populated Asian countries, e-bikes have become a new supernova in daily urban transportation. To facilitate the operations of e-bike-based mobility, the present paper studies the management of the battery deployment for the e-bike battery-swapping system, where the unique features of e-bike riding are considered. Given the pedal-assisted mode, e-bike users could abandon waiting and return to the station later on without too much range anxiety. However, because of the time-varying nature of the customer arrival and the complicated user behaviors, the battery quantity at each station is altered to guarantee the designated service level. However, little research has been done on the operations management of the e-bike battery-swapping system. To bridge the gap, we propose a nonstationary queueing network model to characterize the customer behaviors during the battery-swapping service. Then we develop a closed-form delayed infinite-server fluid approximation for the battery deployment of the one-time-loop scenario under various quality-of-service targets. In addition, we handle the infinite-time-loop scenario with the simulation-based iterative staffing algorithm. In the simulation study, we observe that the proposed battery deployment algorithms can help stabilize the system performance in terms of abandonment probability and expected delay in the face of time-varying demand and complex customer behaviors. Moreover, we reveal that the number of return loops correlates with the service level targets on the battery deployment decision. Furthermore, a time gap exists between the demand and the optimal battery deployment, making proactive battery management in the system possible. Funding: This work was supported by the National Natural Science Foundation of China [Grants 72271222, 71871203, 71872093, 72271137, L1924063], and the National Social Science Fund of China [Grant 21&ZD128]. Supplemental Material: The online appendix is available at https://doi.org/10.1287/trsc.2022.0132 .

TEKNIS OPERASIONAL PENGELOLAAN SAMPAH DI DESA NAGRAK DAN DESA CIANGSANA MENUJU TPST 3R

Nagrak and Ciangsana Village located in Gunung Putri Sub-district, Bogor Regency, West Java. The area of Nagrak Village is 5.84 km2 and the area of Ciangsana Village is 8.62 km2. The level of waste service in both villages is 12%, the planned service level target at the end of the planning year 2032 is 100%. This planning data collection uses the SNI 19-3964-1994 method regarding data collection methods in operational technical planning for waste management. Sampling using the stratified random sampling method, the waste generation rate in Nagrak Village and Ciangsana Village is 2 liters/person/day with a total population of 51488 people from the two villages, the waste generation from domestic areas has 54% organic waste and anorganic waste has 46%. The waste generation from Non-domestic areas has 60% organic waste and anorganic waste has 40%. Organic waste is processed with Black Soldier Fly (BSF), anorganic waste is processed with the Refused Derived Fuel (RDF) method at TPST 3R Trisakti. The planning stage consists of short term (2022-2024) medium term (2025-2028) and long term (2029-2032). The cost budget plan includes investment, operational and maintenance costs with a total cost of Rp.59,424,025,515. The selected alternative is alternative I, which has a more economical and affordable cost budget for Nagrak Village and Ciangsana Village.

The impact of stockout‐based switching on fill rates

AbstractThe possibility of product substitution due to a stockout provides retailers with a buffer against lost sales but can also complicate the management of inventory because both substitute and primary demand affect inventory levels. Since most retailers make their inventory decisions around a desired service level (e.g., fill rate), it is important to understand how customer substitution behavior can influence different measures of customer service. In this study, an extensive theoretical framework is presented and used to develop a decision tree approach for predicting realized item and category fill rates. We find that item fill rate is primarily a function of the target service level of the focal item and the willingness to switch from an alternate item to the focal item. Category fill rate is influenced by the target service level of both items, with willingness to switch amplifying their effect. The decision tree approach is found to be an accurate predictor in most cases, however, it tends to overestimate item fill rate when the willingness to substitute from the alternate item increases. The approach also accurately predicts category fill rates, outside of scenarios with asymmetric substitutability.

A STOCHASTIC APPROACH FOR AGGREGATE PRODUCTION PLANNING CONSIDERING RANDOM DEMAND AND RANDOM QUALITY

ABSTRACT: This paper presents a multi-stage stochastic model to address an aggregate production planning problem. The proposed stochastic model includes two random variables, namely, random demand and random quality, where the production system generates a random amount of defectives. The main contribution of this paper is to extend the traditional deterministic production planning model to take into account the impact of uncertainty, which yields to the development of an integrated stochastic model that jointly optimizes production management decisions considering quality concerns under a certain service level. The performance of the proposed stochastic model is analyzed to draw relevant managerial insights regarding the effects of such a service level target. Also, a sensitivity analysis is conducted to study the effect of varying several costs and system uncertainty parameters to validate the cost-effectiveness of integrated stochastic modelling. The obtained results demonstrate that uncertainty has a significant influence on the decision variables and the total incurred cost, and it is not advisable to ignore the presence of the stochastic variables since they significantly modify the optimal decisions. Keywords: aggregate production planning, quality, random demand, stochastic programming, optimization

The Generalized c/μ Rule for Queues with Heterogeneous Server Pools

Dynamic Routing of Queues with Heterogeneous Server Pools In “The Generalized c/μ Rule for Queues with Heterogeneous Server Pools,” Long, Zhang, Zhang, and Zhang study the optimal control of queueing systems with heterogeneous server pools and a single customer class. The goal is to balance the holding cost of the queue with the operating costs of the server pools. They introduce the target-allocation policy, the Gc/μ rule, and the fixed priority policy for systems with general, convex, and concave cost functions, respectively. They also consider an extension to minimize operating costs and maintain a service-level target for customers waiting in the queue. Moreover, they show that their asymptotically optimal routing policies coincide with several classic policies in the literature in special cases.

Replenishment strategies for lost sales inventory systems of perishables under demand and lead time uncertainty

We develop an inventory control policy for perishable products considering both random demand and random lead time. We consider a B2C retail environment where excess demand is lost. The policy dynamically determines the optimal replenishment quantity under a service level constraint in every period, allowing for order-crossing, a widely disregarded characteristic in the literature. Regarding perishability, we compare the two most extreme issuing policies, first-expired-first-out (FEFO) and last-expired-first-out (LEFO), and evaluate our policy to existing inventory policies for perishables that typically ignore lead time uncertainty.We obtain several interesting findings. First, we show that ignoring lead time uncertainty and planning based on the expected lead time significantly undershoots the target service level. Even planning with the maximum lead time, under LEFO, the achieved service level would still fall considerably below the target, which the lost-sales structure can explain. On the other hand, under FEFO, the achieved service level would overshoot the target service level, which leads to unnecessary waste. Second, a more reliable lead time can significantly reduce waste, especially under LEFO. Third, our model allows us to distinguish between past, present, and future lead time uncertainty and thus to consider partial lead time information. We show the value of lead time information on outstanding orders. Fourth, we evaluate the impact of a fast but unreliable delivery option and a slow but reliable delivery option on the retailer’s average waste and ordering process. We find that the optimal choice depends on the demand characteristics.

Agile roadmap for application‐driven Multi‐UAV networks: The case of COVID‐19

Drones, also known as Unmanned Aerial Vehicles (UAVs), are about to bring drastic transformations to our world and daily lives. News thinking and efficient deployment are required to boost the adoption of UAV‐augmented commercial/civil applications. Yet, network service providers are still facing several design challenges of UAV‐assisted application, due to lack of a roadmap allowing to meet the target service level agreement requirements. In this paper, we propose a complete framework for the UAV as a service paradigm, integrating all the actors/stakeholders contributing to the UAV‐augmented service, and draw their interactions using data/service/money flows. Next, we instantiate our framework on the COVID‐19‐like pandemics, and discuss how to use it force social distancing, spray disinfectants, broadcast messages, deliver medical supplies and enhance surveillance. Computer simulation provide insights on how to set the multi‐UAV network to combat COVID‐19.

Need for speed, but how much does it cost? Unpacking the fee‐speed relationship in Bitcoin transactions

AbstractThe growing popularity of blockchain‐based cryptocurrencies is driven by the flexibility in transaction fee offerings, among other factors. To achieve service‐level differentiation among their users, many cryptocurrencies allow users to “name your own price,” giving rise to a large variation in fee offerings and hence, variation in confirmation times. Yet, the time it takes a cryptocurrency transaction to be confirmed in the blockchain is not only affected by the fee offered, but also by the contemporaneous congestion level and the inherent randomness in the verification process. Although it is generally expected that higher fees lead to quicker confirmation, the uniqueness of the cryptocurrency setting adds important nuances to the fee‐speed relationship. Using Bitcoin—the original and most heavily used cryptocurrency by far—as our empirical context, we stylize the transaction confirmation processes, propose a theoretical framework that maps the causal path from fee to speed, and estimate this framework using Bitcoin transaction data under periods of high volatility. Our results show strong evidence for two characteristics of fee's impact: congestion dependence and tail shrinkage. Our finding that the speed acceleration effect of fee is particularly strong on the tail of the confirmation time distribution motivates a target service level approach to fee recommendation. To put this finding into practice, we develop an efficient computational procedure that helps Bitcoin users accurately estimate fees based on their confirmation delay preferences. We discuss the implications of our analyses on future cryptocurrency development and the long‐term adoption of this revolutionary technology.

Investigating temporal dynamics of urban densification on the buried water infrastructure performance

Urban densification has dominated the land uses and land covers which may jeopardize the quality of life and sustainable development. Moreover, urban densification process exerts stress on buried water infrastructure (BWI) performance. In this study, the need of predicting the change in urban densification and its subsequent impacts on BWI performance over time has drawn attention to conduct a temporal analysis. A hybrid approach was used to select the significant control performance indicators (CPIs) and conduct a temporal analysis. Eight scenarios were generated using the Canadian Council of the Ministers of the Environment Water Quality Index approach to address the change in drinking water quality. The proposed approach was implemented in a municipality of British Columbia, Canada, as a case study. The results showed a 40 % decrease in the BWI performance due to the urban areas have rapidly grown both in population and built-up areas over three decades. Furthermore, an ideal situation was assumed to show that the targeted service level (i.e., 0.5) can be achieved with 3 % to 5 % improvement in every five years of service level. These findings will assist the decision makers to understand the linkage between CPIs to maintain and achieve the required BWI performance.

The data-driven newsvendor problem: Achieving on-target service-levels using distributionally robust chance-constrained optimization

The classical approach to the newsvendor problem is to first estimate the demand distribution (or assume it to be given) and then determine the optimal inventory level. Data-driven optimization offers an alternative, where the inventory level is determined directly from the data. In this paper, we consider the data-driven newsvendor problem under a service-level constraint. We show that existing approaches to this problem suffer from overfitting, resulting in service-levels that are below the target service-level. We propose new data-driven approaches and corresponding mathematical optimization models based on methods of distributionally robust chance-constrained optimization—which have not yet been applied and empirically tested in the context of the data-driven newsvendor problem. We assess the effectiveness of our approaches by means of an extensive numerical study. To that end, we conduct structured experiments based on simulation as well as experiments based on a real-life bikesharing system where we consider the daily usage data along with information on weather and seasonal factors. The results demonstrate that our methods achieve on-target service-levels even in absence of large amounts of data. All in all, our study provides ample empirical evidence that distributionally robust chance-constrained optimization is a viable approach for addressing the data-driven newsvendor problem.

Empirical newsvendor biases: Are target service levels achieved effectively and efficiently?

Human decision making in the newsvendor context has been analyzed intensively in laboratory experiments, where various decision biases have been identified. However, it is unclear whether the biases also exist in practice. We analyze the ordering decisions of a manufacturer who faces a multiproduct newsvendor problem with an aggregate service‐level constraint. We find that the manufacturer broadly exhibits the same biases as subjects in the laboratory and is prone to another bias that has not been identified before, that is, group aggregation. The bias can be attributed to the multi‐product problem of the manufacturer, and refers to the observation that the service levels are not optimized for individual products, but rather for product groups. Our data allow us to analyze the performance of a manufacturer in detail and we find that target service levels are achieved effectively, but not efficiently. We provide rationales for the manufacturer's ordering behavior, discuss managerial implications, and quantify the financial benefits of debiasing ordering decisions.

Timing intermittent demand with time-varying order-up-to levels

Current intermittent demand inventory control models assume that the demand interval is memoryless: the probability of observing a positive demand does not depend on the time since the last demand occurred. Contrarily, several forecasting contributions suggest that demand intervals contain more distributional information. We find that the data of the M5 forecasting competition confirms this. Therefore, we propose an inventory control model that explicitly uses the full distributions of the demand sizes and intervals and thereby acknowledges that the probability of a demand occurrence may vary throughout the interval. To exploit this information, we also allow for time-varying order-up-to levels that flexibly adjust inventories according to the dynamic requirements. We derive the long-run average holding costs, non-stockout probability, order fill rate, and volume fill rate. Inspired by an analogy with multi-item inventory control models, we propose a greedy marginal-analysis heuristic to optimize the order-up-to levels, which we benchmark against the optimal solution on theoretical instances. In a simulation study on the M5 competition data we demonstrate this method’s improved on-target service performance compared to that of traditional solutions. We furthermore show that target service levels can be achieved at significantly lower costs with time-varying than with fixed order-up-to levels.

An empirical Bayes approach to incorporating demand intermittency and irregularity into inventory control

Spare parts inventory management is complex due to the combined impact of intermittent and variable demand patterns. It becomes even more challenging if the spare parts demand distribution is highly complex due to strong interdependent demand intermittency and extremely irregular demand. The research literature proposes many analytical methods for forecasting spare parts demand. But, due to their limited flexibility in modeling complex demand patterns, existing forecasting methods may not produce satisfactory results for a spare parts portfolio displaying extremely complex demand patterns. This study proposes a novel nonparametric Bayesian forecasting approach with its roots in the empirical Bayes paradigm. The method is subject to few performance constraints and is highly flexible in dealing with a rich diversity of demand patterns, including extreme demand complexity. We assess the relative performance of this new approach with several prominent methods in the literature using an automotive parts distributor's empirical demand data for 46,272 stock-keeping units. This dataset is representative of typical spare parts portfolios that are characterized by a wide variety and extremely complex demand patterns. The experimental findings show the new Bayesian approach achieves the best overall performance in terms of inventory efficiency and minimal backorders for meeting specified target service levels. This favorable performance reflects the approach's flexibility to accommodate disparate and complex demand patterns, including interdependence of demand intermittency, irregular demand distribution, and even nonstationary demand distribution to some extent, and provide robust solutions.