In today’s fast-moving world, a company´s ability to align with changes in the market is becoming a major competitive factor. Demand forecasting form the basis of all supply chain planning and is a process that companies often fail to recognize as a key contributor to corporate success. Different contexts and market dynamics creates different challenges for companies to overcome in order to have an efficient forecasting process, matching demand with supply. This master thesis looks at the whole forecasting process, also called forecasting management, at a decentralized global company to identify the main challenges within the process and propose recommendations on how to overcome them. The research is based on a single case study where the forecasting process is investigated using four different dimensions: Functional Integration, Approach, Systems and Performance Measurements. The study identified twelve challenges in the forecasting process where a majority can be connected to issues within information sharing and lack of support in the process. Based on the identified challenges, eight improvement suggestions where developed to target the challenges and improving the process for a decentralized global company. / I dagens snabbt utvecklande och växande landskap så är ett företags förmåga att anpassa sig till marknadens behov en betydande konkurrensfaktor. Säljprognoser utgör grunden för all planering inom försörjningskedjan och är en process som företag ofta inte erkänner som en viktig bidragsgivare till företagets framgång. Olika marknadslandskap och förutsättningar skapar olika utmaningar för företag att bemästra för att kunna bedriva ett effektivt prognosarbete och matcha efterfrågan med utbud. Detta examensarbete tittar på hela prognosprocessen, även kallad prognoshantering, hos ett decentraliserat globalt företag för att identifiera de viktigaste utmaningarna i processen och föreslå rekommendationer om hur man kan övervinna dem. Forskningen bygger på en enda fallstudie där prognosprocessen undersöks utifrån fyra olika dimensioner: Funktionell integration, strategi, system och prestandamätningar. Studien identifierade tolv utmaningar i prognosprocessen där en majoritet kan kopplas till utmaningar inom informationsdelning och brist på stöd i processen. Baserat på de identifierade utmaningarna utvecklades åtta förbättringsåtgärder för att övervinna utmaningarna och förbättra processen för ett decentraliserat globalt företag.

This study aims to enhance the accuracy of a demand forecasting model, XGBoost, by incorporating synthetic multivariate restaurant time series data during the training process. The research addresses the limited availability of training data by generating synthetic data using TimeGAN, a generative adversarial deep neural network tailored for time series data. A one-year daily time series dataset, comprising numerical and categorical features based on a real restaurant's sales history, supplemented by relevant external data, serves as the original data. TimeGAN learns from this dataset to create synthetic data that closely resembles the original data in terms of temporal and distributional dynamics. Statistical and visual analyses demonstrate a strong similarity between the synthetic and original data. To evaluate the usefulness of the synthetic data, an experiment is conducted where varying lengths of synthetic data are iteratively combined with the one-year real dataset. Each iteration involves retraining the XGBoost model and assessing its accuracy for a one-week forecast using the Root Mean Square Error (RMSE). The results indicate that incorporating 6 years of synthetic data improves the model's performance by 65%. The hyperparameter configurations suggest that deeper tree structures benefit the XGBoost model when synthetic data is added. Furthermore, the model exhibits improved feature selection with an increased amount of training data. This study demonstrates that incorporating synthetic data closely resembling the original data can effectively enhance the accuracy of predictive models, particularly when training data is limited.

Demand forecasting is a well-established internal process at Ericsson, where employees from various departments within the company collaborate in order to predict future sales volumes of specific products over horizons ranging from months to a few years. This study aims to evaluate current predictions regarding radio unit products of Ericsson, draw insights from historical volume data, and finally develop a novel, statistical prediction approach. Specifically, a two-part statistical model with a decision tree followed by a neural network is trained on previous sales data of radio units, and then evaluated (also on historical data) regarding predictive accuracy. To test the hypothesis that mid-range volume predictions of a 1-3 year horizon made by data-driven statistical models can be more accurate, the two-part model makes predictions per individual radio unit product based on several predictive attributes, mainly historical volume data and information relating to geography, country and customer trends. The majority of wMAPEs per product from the predictive model were shown to be less than 5% for the three different prediction horizons, which can be compared to global wMAPEs from Ericsson's existing long range forecast process of 9% for 1 year, 13% for 2 years and 22% for 3 years. These results suggest the strength of the data-driven predictive model. However, care must be taken when comparing the two error measures and one must take into account the large variances of wMAPEs from the predictive model. / Ericsson har en väletablerad intern process för prognostisering av försäljningsvolymer, där produktnära samt kundnära roller samarbetar med inköpsorganisationen för att säkra noggranna uppskattningar angående framtidens efterfrågan. Syftet med denna studie är att evaluera tidigare prognoser, och sedan utveckla en ny prediktiv, statistisk modell som prognostiserar baserad på historisk data. Studien fokuserar på produktkategorin radio, och utvecklar en två-stegsmodell bestående av en trädmodell och ett neuralt nätverk. För att testa hypotesen att en 1-3 års prognos för en produkt kan göras mer noggran med en datadriven modell, tränas modellen på attribut kopplat till produkten, till exempel historiska volymer för produkten, och volymtrender inom produktens marknadsområden och kundgrupper. Detta resulterade i flera prognoser på olika tidshorisonter, nämligen 1-12 månader, 13-24 månader samt 25-36 månder. Majoriteten av wMAPE-felen för dess prognoser visades ligga under 5%, vilket kan jämföras med wMAPE på 9% för Ericssons befintliga 1-årsprognoser, 13% för 2-årsprognerna samt 22% för 3-årsprognoserna. Detta pekar på att datadrivna, statistiska metoder kan användas för att producera gedigna prognoser för framtida försäljningsvolymer, men hänsyn bör tas till jämförelsen mellan de kvalitativa uppskattningarna och de statistiska prognoserna, samt de höga varianserna i felen.

No description available.

No description available.

Finding the proper balance between availability and cost efficiency is an important concern within spare part management. Spare part suppliers need to respond quickly to customer demand as a stock-out can have severe consequences for both the customer and the supplier. It is critical to identify what items to keep in stock and where to allocate the inventory to avoid stock-outs. This case study was performed at a large high-tech company producing manufacturing equipment to be used in the electronics industry. The aim was to lower the stock-levels of spare parts while not impairing the availability by combining item classification, demand forecasting, and distribution network optimization. A decision diagram for classifying spare parts was constructed using the analytical hierarchy process. Twenty items were classified using the diagram, and the demand for them was forecasted using the Syntetos Boylan Approximationmethod. The shipping cost for spare parts within one region was minimized using a linear optimization model. The analysis showed that equipment criticality, annual usage value, and installed base are critical when classing spare parts. Instead of using five distribution centers in the European region, it was discovered that the shipping costs would decrease if only three warehouses made up the distribution network. The spare parts investigated appeared to follow the typical characteristics for spare parts, showing a low and irregular demand. Hence, demand forecasting seemed to be unnecessary, considering the difficulties in getting satisfactory results. Instead of combining the results from classification, forecasting, and inventory allocation, we suggest that the processes affecting stocking decisions should cooperate and work towards a common objective, namely to satisfy the customer demand in a cost-efficient way. Thus, widening the meaning of taking on an integrated approach to spare part management. / Inom hanteringen av reservdelar är det en stor utmaning att hitta rätt avvägning mellan tillgänglighet och kostnadseffektivitet. Leverantörer av reservdelar måste snabbt kunna möta kundefterfrågan eftersom uteblivna leveranser av kritiska reservdelar kan få allvarliga konsekvenser för både kund och leverantör. Vilka artiklar som ska lager-hållas och var de ska lagerhållas är avgörande beslut för att undvika att artiklar rest-noteras. I den här fallstudien, som utfördes på ett stort teknikföretag som tillverkarproduktionsutrustning till elektronikindustrin, var syftet att sänka lagernivåerna av reservdelar utan att göra avkall på tillgängligheten. Detta genom att kombineragruppering av artiklar, beräkning av kommande efterfrågan och optimering av distributionsnätverket. För att klassificera artiklar i grupper med liknande egenskaper skapades ett schematiskt beslutsdiagram med hjälp av metoden AHP. Tjugo artiklar ur sortimentet valdes ut som beslutsdiagrammet testades på. För samma tjugo artiklar gjordes prognoser för den kommande efterfrågan med metoden Syntetos-Boylan-Approximation. Distributionsnätverket i den europeiska regionen optimerades medavseende på fraktkostnad genom att applicera en linjär optimeringsmodell. Hur kritisk en reservdel är för den relaterade maskinens funktionalitet, reservdelensårliga förbrukningsvärde och den geografiska placeringen av installerade maskinervisade sig vara kritiska för att kunna klassificera artiklarna effektivt. Analysen av distributionsnätverket i Europa visade att fraktkostnaderna kan minskas om nätverket utgjordes av tre lager istället för fem som det gör i dagsläget. De tjugo undersökta reservdelarna uppvisade de typiska egenskaperna för reservdelar som har rapporterats i litteraturen som låg och oregelbunden efterfrågan. Att sätta prognoser på efterfrågan verkar obefogat med tanke på komplexiteten i beräkningarna och att de ger få tillfredsställande resultat. Istället för att kombinera resultaten från klassificering, prognoser på efterfrågan och lageroptimering föreslår vi att alla de funktioner i ett företag som arbetar med att tillgodose kundefterfrågan bör samarbeta i högre grad och jobba mot ett gemensamt mål, nämligen att tillgodose kundernas efterfrågan på ett kostnadseffektivt sätt. Således vill vi utvidga betydelsen av att ta en integrerad strategi för reservdelshantering

Detta examensarbete undersöker användningen av datadrivna metoder för efterfrågan prognoser och lagerstyrning inom ramen för Ericssons supply chain management. Studien fokuserar på integrationen av maskininlärning, demand shaping och realtidsdata för att förbättra noggrannheten och effektiviteten inom dessa avgörande områden. Studien utforskar effekten av maskininlärningstekniker på efterfråganprognoser och betonar betydelsen av exakta förutsägelser för att vägleda produktion, lagerhantering och distributionsstrategier. För att implementera detta föreslår studien integrationen av realtidsdataströmmar och Internet of Things (IoT)-enheter, vilket möjliggör insamling av aktuell information. Denna integration underlättar snabba svar på varierande efterfrågemönster och optimerar därmed supply chain-operationer. Studien ger värdefulla insikter för Ericsson för att förbättra sina förmågor inom efterfråganprognoser och för att optimera lagerhanteringen i en datadriven miljö. / This master's thesis investigates the utilization of data-driven approaches for demand forecasting and inventory control in the context of Ericsson's supply chain management. The study focuses on the integration of machine learning, demand shaping, and real-time data to enhance accuracy and efficiency in these critical areas. The research explores the impact of machine learning techniques on demand forecasting, highlighting the significance of precise predictions in guiding production, inventory management, and distribution strategies. To address this, the study proposes the integration of real-time data streams and Internet of Things (IoT) devices, enabling the capture of up-to-date information. This integration facilitates prompt responses to evolving demand patterns, thereby optimizing supply chain operations.The research provides valuable insights for Ericsson to enhance its demand forecasting capabilities and optimize inventory management in a data-driven environment.

Poor input data quality used in repairable spare parts forecasting by aerospace small and midsize enterprises (SME) suppliers results in poor inventory practices that manifest into higher costs and critical supply shortage risks. Guided by the data quality management (DQM) theory as the conceptual framework, the purpose of this exploratory multiple case study was to identify the key strategies that the aerospace SME repairable spares suppliers use to maximize their input data quality used in forecasting repairable spare parts. The multiple case study comprised of a census sample of 6 forecasting business leaders from aerospace SME repairable spares suppliers located in the states of Florida and Kansas. The sample was collected via semistructured interviews and supporting documentation from the consenting participants and organizational websites. Eight core themes emanated from the application of the content data analysis process coupled with methodological triangulation. These themes were labeled as establish data governance, identify quality forecast input data sources, develop a sustainable relationship and collaboration with customers and vendors, utilize a strategic data quality system, conduct continuous input data quality analysis, identify input data quality measures, incorporate continuous improvement initiatives, and engage in data quality training and education. Of the 8 core themes, 6 aligned to the DQM theory's conceptual constructs while 2 surfaced as outliers. The key implication of the research toward positive social change may include the increased situational awareness for SME forecasting business leaders to focus on enhancing business practices for input data quality to forecast repairable spare parts to attain sustainable profits.

Turkey, with its climate, protected bays, cultural and environmental resources is an ideal place for yacht tourism. Subsequently, yacht tourism is increasing consistently. Yacht tourism can cause unmitigated development and environmental concerns when aiming to achieve tourist satisfaction. As the demand for yacht tourism intensifies, sustainable development strategies are needed to maximize natural, cultural and economic benefits. Integration of forecasts to the strategic planning is necessary for sustainable and use of the coastal resources. In this study two different quantitative forecasting techniques - Exponential smoothing and Auto-Regressive Integrated Moving Average (ARIMA) methods were used to estimate the demand for yacht berthing capacity demand till 2030 in Turkey. Based on environmental, socio-economic and geographic data and the opinions gathered from stakeholders such as marina operators, local communities and government officials an allocation model was developed for the successful allocation of the predicted demand seeking social and economical growth while preserving the coastal environment. AHP was used to identify and evaluate the development, social and environmental and geographic priorities. Aiming a dynamic plan which is responsive to both national and international developments in yacht tourism, potential investment areas were determined for the investments required to accommodate the future demand. This study provides a multi dimensioned point of view to planning problem and highlights the need for sustainable and dynamic planning at delicate and high demand areas such as coasts.

Demand forecasting performance is subject to the uncertainty underlying the time series an organisation is dealing with. There are many approaches that may be used to reduce demand uncertainty and consequently improve the forecasting (and inventory control) performance. An intuitively appealing such approach that is known to be effective is demand aggregation. One approach is to aggregate demand in lower-frequency 'time buckets'. Such an approach is often referred to, in the academic literature, as temporal aggregation. Another approach discussed in the literature is that associated with cross-sectional aggregation, which involves aggregating different time series to obtain higher level forecasts.This research discusses whether it is appropriate to use the original (not aggregated) data to generate a forecast or one should rather aggregate data first and then generate a forecast. This Ph.D. thesis reveals the conditions under which each approach leads to a superior performance as judged based on forecast accuracy. Throughout this work, it is assumed that the underlying structure of the demand time series follows an AutoRegressive Integrated Moving Average (ARIMA) process.In the first part of our1 research, the effect of temporal aggregation on demand forecasting is analysed. It is assumed that the non-aggregate demand follows an autoregressive moving average process of order one, ARMA(1,1). Additionally, the associated special cases of a first-order autoregressive process, AR(1) and a moving average process of order one, MA(1) are also considered, and a Single Exponential Smoothing (SES) procedure is used to forecast demand. These demand processes are often encountered in practice and SES is one of the standard estimators used in industry. Theoretical Mean Squared Error expressions are derived for the aggregate and the non-aggregate demand in order to contrast the relevant forecasting performances. The theoretical analysis is validated by an extensive numerical investigation and experimentation with an empirical dataset. The results indicate that performance improvements achieved through the aggregation approach are a function of the aggregation level, the smoothing constant value used for SES and the process parameters.In the second part of our research, the effect of cross-sectional aggregation on demand forecasting is evaluated. More specifically, the relative effectiveness of top-down (TD) and bottom-up (BU) approaches are compared for forecasting the aggregate and sub-aggregate demands. It is assumed that that the sub-aggregate demand follows either a ARMA(1,1) or a non-stationary Integrated Moving Average process of order one, IMA(1,1) and a SES procedure is used to extrapolate future requirements. Such demand processes are often encountered in practice and, as discussed above, SES is one of the standard estimators used in industry (in addition to being the optimal estimator for an IMA(1) process). Theoretical Mean Squared Errors are derived for the BU and TD approach in order to contrast the relevant forecasting performances. The theoretical analysis is supported by an extensive numerical investigation at both the aggregate and sub-aggregate levels in addition to empirically validating our findings on a real dataset from a European superstore. The results show that the superiority of each approach is a function of the series autocorrelation, the cross-correlation between series and the comparison level.Finally, for both parts of the research, valuable insights are offered to practitioners and an agenda for further research in this area is provided.