Infrequent Demand Research Articles

Bipropellant Enabled Solid Oxide Fuel Cell Stack Power Supply: Energy Agility for Space Vehicles

Spacecraft often have instances where they are required to operate for long periods of time with significant power requirements and no access to solar power. While a battery system could be designed to meet these occasional, and possibly unexpected requirements, creating a power supply to meet these infrequent demands results in having a battery system that is excessively large, massive, and inefficient. A Solid Oxide Fuel Cell (SOFC) system capable of running on the same fuel and oxidizer as the onboard thruster of a bipropellant thruster equipped spacecraft can eliminate much of the added mass and volume needed to enable batteries to meet extreme power demands and create more operational flexibility on how the chemical energy in stowed bipropellants are used. If the spacecraft needs thrust, the bipropellant can be ignited to provide thrust. If the spacecraft needs electric power, the spacecraft can consume bipropellant to generate auxiliary power. In this work, the proven bipropellant pair ammonia (NH3) and nitrous oxide (N2O) are shown to be a viable fuel and oxidant for hydrogen (H2) consuming air breathing SOFCs, as spontaneous NH3 decomposition will generate H2 and spontaneous N2O decomposition will generate O2. A scale up to a 5-cell SOFC stack fed NH3 and N2O produced power up to ~89W. Higher power can be achieved by scaling up the number of cells in a stack to achieve required power demands. The results of this study demonstrate the potential and feasibility of integrating SOFCs with a bipropellant thruster to creating a dual functioning spacecraft system that can satisfy operational power requirements and on-orbit maneuverability.This material is based on research sponsored by Air Force Research Laboratory under agreement number FA9453-21-C-0060. The views expressed are those of the authors and do not reflect the official guidance or position of the United States Government, the Department of Defense or of the United States Air Force. Approved for public release; distribution is unlimited. Public Affairs release approval #AFRL20230023. Figure 1

The Influence of Repair Quality on Aircraft Spare Part Demand Variability

Accurate estimation of spare part demand is challenging in the case of intermittent or lumpy demand, characterised by infrequent demand occurrence and variability in demand size. While prior research has considered the effect of exogenous variables on spare part demand, there is a lack of research considering the effects of repair quality and aggregated spare part demand behaviour across fleets of assets under the influence of multiple simultaneously acting drivers of failure. This research provides new insights towards the problem of estimating variable spare part demand through modelling and simulation of the effects of multiple, simultaneously considered spare part demand drivers. In particular, a contribution to the state of the art is introduced by the use of a Branching Poisson Process (BPP) to model repair quality effects for spare part demand generation in conjunction with several demand drivers. The approach is applied in a numerical study which involves component failure characteristics based on real-life data from an aircraft maintenance, repair and overhaul (MRO) provider. It is shown that repair quality improvements drive down the variance in the demand and the total number of failures over time, and outperform the effect of environmental drivers of failure in terms of demand generation.

Combined Forecasts of Intermittent Demand for Stock-keeping Units (SKUs)

Effective inventory management requires accurate forecasts for stock-keeping units (SKUs), especially for the strategic ones for companies’ operations and after-sales services like providing spare parts. Forecasting is a challenging task for such SKUs as they usually have intermittent demand (ID) patterns, consisting of many periods with zero demand and infrequent demand arrivals. Given the highly uncertain nature of ID for SKUs, this study developed a methodological framework for combining statistical and judgmental forecasts and assessed the performance of the proposed framework by using accuracy and bias measures. The forecasting process has several steps, including data preparation, data categorization based on demand patterns, generating statistical and judgmental forecasts, combining statistical and judgmental forecasts, and evaluating the forecast performance. These steps were illustrated on a real-world dataset that contains monthly customer demand data for after-sales spare parts. Results showed that combination is the best method for the majority of SKUs. This paper contributes to the limited literature by addressing the gap between the combined and ID forecasts. The proposed framework gives practitioners and researchers a comprehensive overview to help them make more accurate forecasts while encouraging the use of simple but structured approaches.

Intermittent Demand Forecasting Using LSTM With Single and Multiple Aggregation

Intermittent demand data is data with infrequent demand with varying number of demand sizes. Intermittent demand forecasting is useful for providing inventory control decisions. It is very important to produce accurate forecasts. Based on previous research, deep learning models, especially MLP and RNN-based architectures, have not been able to provide better intermittent data forecasting results compared to traditional methods. This research will focus on analyzing the results of intermittent data forecasting using deep learning with several levels of aggregation and a combination of several levels of aggregation. In this research, the LSTM model is implemented into two traditional models that use aggregation techniques and are specifically used for intermittent data forecasting, namely ADIDA and MAPA. The result, based on tests on the six predetermined data, the LSTM model with aggregation and disaggregation is able to provide better test results than the LSTM model without aggregation and disaggregation.

A Novel Approach of Intermittent Demand Prediction in Industrial Domain

Intermittent demand, one of the categories of unusual demand, such as demand of spare parts of industry equipment management, seasonal and short life cycle products, is characterized by infrequent demand arrivals and variable demand sizes, which is difficult to predict. To solve this problem, a novel approach was proposed in this paper to predict intermittent demand. It provided mechanism to forecast demand arrival time with demand values at the same time when demand occurs. It firstly applied Decision Tree to predict a 0-1 binary values of demand arrival time. Meanwhile Neural Network was built to predict demand values. Then the time points of these two results were matched. The demand values for those moments when predicted demand that did not occur were replaced by zero. Finally we applied this approach to predict sparse demand of products, the results showed that the predicting accuracy was superior to traditional Back-propagation Neural Network through comparison.

Demand cycles and market segmentation in bicycle sharing

Consumers often display unique habitual behaviors, and knowledge of these behaviors is of great value in prediction of future demand. We investigated consumer behavior in bicycle sharing in Beijing, where demand prediction is critical for cost-effective rebalancing of bicycle locations (putting bikes where and when they will be rented) and supply (number of bicycles). We created baseline statistical demand models, borrowing methods from economics, signal processing and animal tracking to find consumption cycles of 7, 12, 24 h and 7-days. Lorenz curves of bicycle demand revealed significant stratification of consumer behavior and a long-tail of infrequent demand. To overcome the limits of traditional statistical models, we developed a deep-learning model to incorporate (1) weather and air quality, (2) time-series of demand, and (3) geographical location of demand. Customer segmentation was added at a later stage, to explore potential for improvement with customer demographics. Our final machine learning model with tuned hyperparameters yielded around 50% improvement in predictions over a discrete wavelet transform model, and 80–90% improvement in predictions over a naïve model the reflects some current industry practice. We assessed causality in the deep-learning model, finding that location and air quality had the strongest causal impact on demand. The extreme market segmentation of customer demand, and our relatively short time span of data combined to make it difficult to find sufficient data on all customers for a model fit based on segmentation. We reduced our model data to only the 10 most frequent to see whether such segmentation improves our model's predictive success. These results, though limited, suggest that customer behavior within market segments is more stable than across all customers, as was expected.

Demand-Based Computation Offloading Framework for Mobile Devices

Mobile cloud computing is an emerging commercial infrastructure paradigm to enhance the power of smart mobile devices. However, there are significant problems pertaining to efficient provisioning and delivery of mobile applications using cloud IT resources. These barriers concern various levels such as enabling individual mobile user's to dynamically adjust their computing demands and managing these infrequent demands under intermittent network connectivity along with maintaining quality of service (QoS), whereas cloud resources are available for long-time quanta granularity and incur significant setup times. This paper proposes a demand-based computation offloading framework (DBCOF) to balance these competing goals. It provides a mechanism to solve the key problem of application partitioning according to user's demand. In addition, it also reduces the monetary costs per request for the service providers and minimizes service-level agreement (SLA) violations for maintaining QoS. Previous approaches to these problems have focused only on maximizing energy savings in all contexts, irrespective of user's computing requirements. We have implemented the DBCOF prototype for Android and explored its design space by extending CloudSim, the state-of-the-art cloud simulator. Analyses of the results show the effectiveness of DBCOF system in terms of managing computation offloading demands. It also reduces the cost of providing cloud resources to mobile devices to 44.79% and minimizes SLA violations to nearly 1.34% for maintaining the QoS.

PNG: Effective Inventory Control for Items with Highly Variable Demand

LMI developed the PNG inventory control solution to manage inventory items with infrequent demand (i.e., isolated spikes in demand) as well as items with frequent, highly variable demand. Such items account for the majority of hardware stocked at the U.S. Defense Logistics Agency (DLA). The forecasting of demand for these items—no matter how sophisticated the forecasting method—had resulted in years of problems for DLA: excess inventory for some items, backorders for others, and excessive buyer workload. The implementation of PNG, a software package that consists of two inventory solutions, Peak Policy and Next Gen, allowed DLA to shift from trying to forecast each item individually to using a portfolio or risk-management approach to inventory control. Since DLA implemented PNG in January 2013, the agency has achieved its inventory-related goals for better customer service and reduced buyer workload, has experienced no inventory increase, and has saved nearly $400 million per year.

Forecast combinations for intermittent demand

Intermittent demand is characterised by infrequent demand arrivals, where many periods have zero demand, coupled with varied demand sizes. The dual source of variation renders forecasting for intermittent demand a very challenging task. Many researchers have focused on the development of specialised methods for intermittent demand. However, apart from a case study on hierarchical forecasting, the effects of combining, which is a standard practice for regular demand, have not been investigated. This paper empirically explores the efficiency of forecast combinations in the intermittent demand context. We examine both method and temporal combinations of forecasts. The first are based on combinations of different methods on the same time series, while the latter use combinations of forecasts produced on different views of the time series, based on temporal aggregation. Temporal combinations of single or multiple methods are investigated, leading to a new time-series classification, which leads to model selection and combination. Results suggest that appropriate combinations lead to improved forecasting performance over single methods, as well as simplifying the forecasting process by limiting the need for manual selection of methods or hyper-parameters of good performing benchmarks. This has direct implications for intermittent demand forecasting in practice.

A Stochastic Methodology for the Optimal Management of Infrequent Demand Spare Parts in the Automotive Industry

The paper deals with the management of the spare parts in the automotive sector introducing a first definition of business model aimed to reduce holding costs of stocks, focusing on spare parts distributors point of view, and adopting a probabilistic mathematical approach. The management of spare parts is not trivial due to the infrequent profile that characterizes market demand, in contrast with the necessity of the distributors to keep in stock a large number of items (part numbers) in order to avoid lost sales (back ordering is not allowed). Moreover, given that different distributors can deal the same products among the national market, the total amount of stocks of the same products is multiplied along the supply chain. A mathematical approach adopting probabilistic dynamic programming is presented aimed to optimize the management of spare parts, considering the creation of a platform in which different distributors can participate to share the management of items. The main purpose of the model is to find the optimal allocation of products among distributors and, at the same time, ensuring a balancing in the costs that each player has to bear period by period, providing the optimal reorder policy for each product. A numerical example is also reported.

A feasible MapReduce peer-to-peer framework for distributed computing applications

This article presents a MapReduce Peer-to-Peer (P2P) framework that enables a MapReduce implementation on P2P networks to support a class of MapReduce-based computing applications. This framework can be useful for researchers who cannot afford expensive and dedicated clusters for infrequent demands of solving distributed computing problems. The framework also allows Internet users from social and P2P network communities to perform large data processing experiments on distributed environment. The article describes the architecture and prototyping implementation of a MapReduce P2P system. The main features of this system include exploiting leisure computing resources on P2P networks efficiently for computation, providing MapReduce operations using task management for various distributed computing problems, and supporting peer failure management for improving fault tolerance on Internet environment. We have evaluated and compared the MapReduce P2P implementation with a Hadoop MapReduce implementation on local and global-scale networks. The article also includes a discussion of applying the framework to a realistic distributed case-based reasoning system.

More grip on inventory control through improved forecasting: A comparative study at three companies

Inventory control for parts with infrequent demands is difficult since forecasting their demand is problematic. Traditional forecasting methods, such as moving average and single exponential smoothing, are known not to suffice since they do not cope well with periods with zero demands. Croston type methods and bootstrapping methods are more promising. We propose a new bootstrapping method, which we term empirical plus. The added value of this method lies in the fact that it explicitly takes into account that besides the demand, also the supply lead time is stochastic. We compare its performance with a number of methods from all three above-mentioned categories. Opposite to what is done in most comparative studies, we do not focus on performance metrics that are related directly to the forecasting results (e.g., mean squared error), but we focus on the resulting inventory control policy (achieved fill rate and holding costs). We use in our study large data sets from three companies, which we make publicly available. We find that our empirical plus method outperforms the other methods when the average inter-demand interval is large and the squared coefficient of variation of the demand size is small. This class of parts often consists of the expensive parts, for which forecasting is both difficult, because of the infrequent demands, and important, because of the price. The Syntetos Boylan approximation performs best on the other classes of parts. These findings may be used in practice to use the right forecasting method for each type of part, thus achieving more cost-effective spare parts inventory control.

A stochastic programming approach to determine robust delivery profiles in area forwarding inbound logistics networks

One technique to coordinate the suppliers' and the producers' production plans in a supply chain is the use of delivery profiles, which provide fixed delivery frequencies for all suppliers. The selection of a delivery profile assignment has major effects on the cost efficiency and the robustness of a supply chain and thus should be performed carefully. In this work, we consider planning approaches to select delivery profiles for the case of area forwarding-based inbound logistics networks, which are commonly used in several industries to consolidate supplies in an early stage of transport. We present a two-stage stochastic mixed integer linear programming model to determine robust delivery profile assignments under uncertain and infrequent demands and complex tariff systems. The model is embedded into a solution framework consisting of scenario generation and reduction techniques, a decomposition approach, a genetic algorithm, and a standard MILP solver. On the basis of an industrial case study, we show that our approach is computationally feasible and that the planning solutions obtained by our model outperform both a deterministic approach and the planning methodology prevailing in industrial practice.

Forecasting Non-normal Demand by Support Vector Machines with Ensemble Empirical Mode Decomposition

Non-Normal demand is the demand with infrequent demand occurrences or irregular demand sizes, which is very difficult to forecast. In this study, an ensemble empirical mode decomposition (EEMD) based support vector machines (SVMs) learning approach is proposed to forecast demand in these two cases. This approach is under a “decomposition-and-ensemble” principal to decompose the original non-normal demand series into several independent “smooth” and “continuous” subseries including a small number of intrinsic mode functions (IMFs) and a residue by EEMD technique. Then SVMs are used to model each of the subseries so as to achieve more accurate forecast respectively. Finally, the forecasts of all subseries are aggregated by a SVMs model to formulate an ensemble forecast for the non-normal demand series. Two data sets each has four artificial data series were used to test the effectiveness of the proposed approach. Empirical results demonstrate that the proposed ensemble learning approach outperforms the other forecasting methods such as SVMs , ARIMA and Croston method in terms of RMSE, MAPE, MdRAE and MASE.

An aggregate–disaggregate intermittent demand approach (ADIDA) to forecasting: an empirical proposition and analysis

Intermittent demand patterns are characterised by infrequent demand arrivals coupled with variable demand sizes. Such patterns prevail in many industrial applications, including IT, automotive, aerospace and military. An intuitively appealing strategy to deal with such patterns from a forecasting perspective is to aggregate demand in lower-frequency ‘time buckets’ thereby reducing the presence of zero observations. However, such aggregation may result in losing useful information, as the frequency of observations is reduced. In this paper, we explore the effects of aggregation by investigating 5000 stock keeping units from the Royal Air Force (UK). We are also concerned with the empirical determination of an optimum aggregation level as well as the effects of aggregating demand in time buckets that equal the lead-time length (plus review period). This part of the analysis is of direct relevance to a (periodic) inventory management setting where such cumulative lead-time demand estimates are required. Our study allows insights to be gained into the value of aggregation in an intermittent demand context. The paper concludes with an agenda for further research.

Controlling multi-indenture repairable inventories of multiple aircraft parts

This paper presents a model for determining stock levels of repairable items supporting a fleet of commercial aircraft operated by a transportation company in the Philippines. The items are characterised by infrequent demand, high cost and a hierarchical (or indentured) structure. The system has three re-supply sources of serviceable parts, namely, the in-house repair shop, the out-house repair shops, and the suppliers. Non-repairable items are scrapped and replaced with new items on a one-for-one basis. The model considers two levels of indenture represented by modules and components. The objective is to minimise the total expected steady-state annual cost of holding inventories and of aircraft delays. A minimum requirement on module availability is also applied. The formulation is such that the regular discrete optimisers available could not be readily used to solve it. The model is implemented on an illustrative problem, employing an integer search for the item stock levels within a limited range.

Moving Crowds in Chicago: Baseball and the Fourth of July

The Chicago Transit Authority (CTA) is often called on to transport large numbers of spectators attending popular special events. Although these situations bear similarities to normal peak-hour traffic, they are characterized by longer-lasting streams of intense traffic volume. Facilities and operating practices designed for peak-hour traffic are usually unable to handle these extraordinary inflows. Therefore, facilities and operations must be adapted to deal with these extreme conditions. Today, CTA experiences two major types of special-events situations. The first type of situation is sporting events, such as professional baseball, football, and basketball games, which originate predictable traffic from a specific site at predetermined times. Therefore, a nearby transit facility such as a rapid-transit station can be designed to service expected traffic volumes. Second, Chicago hosts a growing list of special events that are located in less-well-defined areas, which may change over time. Attendance at this second type of event is often difficult to predict and measure. Therefore, permanent facilities must be adapted to serve such infrequent demand. Also, special-events operations planning must be coordinated carefully with outside agencies. The discussion focuses on how CTA has designed both facilities and operations to deal with the crowds generated by both scenarios. CTA serves many sporting events, including Cubs baseball games at Wrigley Field. Also, every year Chicago hosts a fireworks display held on the night of July 3. A description is included of the design of facilities and operations that facilitate these special events.

Factors influencing oral health in long term care facilities.

In a stratified random sample of 41 long term care (LTC) facilities in Vancouver, 653 residents were chosen to investigate oral health needs and demands for treatment. All of the 603 dentists in the same area were questioned to assess their interest in attending the residents of the institutions. The information from each source was reviewed to identify factors influencing the oral health services to this predominantly elderly and medically compromised population. The majority (60%) of the residents were edentulous and they made infrequent demands on dentists. Two-thirds of those interviewed said that there was nothing wrong with their mouths, but most of those who were aware of a problem wanted it treated, preferably within the institution. They complained about loose or uncomfortable dentures most frequently, and many were dissatisfied with previous dental treatment. The oral mucosal lesions seen on examination were usually symptomless and associated with poor hygiene, while structurally defective dentures and deep carious lesions were not uncommon. The responding 334 dentists indicated that they enjoyed treating elderly patients, 19% had attended an LTC facility, usually to provide an emergency service, and 37% were willing to provide this service if asked. Interest, however, in the service was curtailed by pressures from private practice, concerns about inadequate training and the small demand and poor conditions in the facilities. Although the demand for treatment was not extensive from the residents, they did have problems that were not receiving care.

PRINCIPAL COMPONENT ANALYSIS OF RATINGS OF SOME DEVIANT ENGLISH SENTENCES

ABSTRACTThe comprehension of deviant sentences, a not infrequent demand in natural situations, is dependent on several linguistic variables. Grammaticalness (G), meaningfulness (M), and familiarity (F) are three variables which are potentially such. In order to study the effect of violating these variables upon Ss' responses to deviant sentences, 85 deviant and 15 correct sentences were assigned to eight groups representing all combinations of two values (“correct” or “deviant”) on these three variables. The 100 sentences were given to four equal groups of Ss (total N = 112), who rated each sentence from 0 to 10 on the basis of either grammaticalness (G*), meaningfulness (M*), familiarity (F*), or ordinariness (O*). The data of the first three groups were then combined into an 84 by 100 matrix. A principal components analysis was performed on the cross‐product matrix with a varimax rotation. Four interpretable factors emerged, accounting for 89% of the variability. Factor I was a general comprehensibility factor in the factor loadings, related to changes in all three variables. However, the familiarity Ss scored highest on Factor I. Factors II and III represented G‐G* and M‐M*, respectively, in both factor loadings and factor scores. Factor IV corresponded to the F variable in the factor loadings, but was uninterpretable for the factor scores.

Diverticulitis of the Colon, with Especial Attention to the Diagnosis

DIVERTICULITIS of the large bowel is of interest to the surgeon because of its ability to simulate other inflammatory conditions in the abdomen, including appendicitis, cholecystitis and acute pelvic inflammatory disease, as well as on account of its not infrequent demand for operative relief or radical cure. It is of interest to the internist because it requires to be differentiated from other causes of vague chronic or recurrent abdominal distress, and because it responds well, in its less acute phases, to medical treatment. It is, however, of even more interest to the roentgenologist, since only he is capable of making a definite diagnosis of this condition in the living patient and since this diagnosis often gives an unexpected and thoroughly satisfying solution to what may have been a puzzling clinical problem. In the recognition of this disease, as in other phases of abdominal X-ray diagnosis, the profession owes a considerable debt to Carman, who was the first roentgenologist to report a series of cases examined by the roentgen ray and the first to formulate a scheme for the recognition and differential diagnosis of diverticulosis. This condition was described from the pathological standpoint by Graser in 1899. In 1907, Mayo, Wilson and Giffin reported five cases which had been treated surgically. The first reported case examined by X-ray was one described by Abbe and examined roentgenologically by LeWald, reported in 1915. Carman's first report was published in 1915 and detailed three cases examined at the Mayo Clinic by X-ray. In the first two of these cases the condition was not recognized from the X-ray films, as Carman had concluded after examination of pathological specimens that it would be impossible to demonstrate diverticula by means of an opaque enema or meal. The characteristic shadows were, however, noted on the films and when the cases subsequently came to operation an extensive multiple diverticulosis was found in both patients. Carman promptly concluded that the small round extralumenal shadows noted must have been due to filled diverticula. The third case of this series had not come to operation at the time of the report, but showed perfectly typical shadows. Since that report, roentgenologists have been finding diverticulosis in an increasing percentage of patients undergoing routine gastro-intestinal examinations, as well as in patients having suggestive symptoms. Case estimates that one in eighty of his routine gastro-intestinal examinations shows diverticula, and our own records show at least as common an incidence as this. As with other more or less rare pathological conditions, the worker who is most zealous in the search for diverticulosis and most open to suggestions of its presence encounters it most frequently. True and false, acquired and congenital, diverticula are described in the literature.