Publications
In the digital economy, information systems have a significant impact on supply chain management. However, there is a need for further development of theoretical knowledge andmathematical models, including methods formanaging risk in complex supply networks to best serve customer orders. In the supply chain operations reference (SCOR) model, reliability is assessed by calculating perfect order parameters. The component/process reliability is calculated as the product of the weighted averages of the perfect order parameters, and possible combinations of failure features are not taken into account. This paper presents an approach to probabilistic estimation of perfect order parameters based on the general theorem on the repetition of experiments, and proposes to use a binomial distribution to approximate the values obtained. The obtained results make it possible to assess the efficiency of possible measures (increasing the insurance stock, replacing the carrier, etc.) to improve the reliability of perfect order fulfilment.
The research into various sources showed that, despite the results achieved, a low-demand environment and short time series are currently often neglected. To improve the reliability and validity of forecast estimates based on a short time series with low demand, it is necessary to create calculation models using all available quantitative and qualitative information. In this paper, we propose an algorithm that includes the systematisation of statistical data in the form of a time series, statistical and analytical models, expert evaluation of forecast consistency, analysis of the results in order to form versions of a combined model for assessing the predicted parameters of stock consumption, making decisions on choosing one of the inventory management strategies for a short time series with low demand, and the proposed approach is tested. Further research of low demand should include a number of directions, in particular, the development of combined forecasting methods, which include, in addition to quantitative and qualitative methods, the application of decision-making methods.
Around 30% to 70% of products in retail and services experience low demand, including spare parts and components for nearly all types of machinery and equipment industries. A detailed analysis of stock forecasting methods for the low demand represents a research gap in inventory management. The existing clustering methods, that is, ABC analysis and XYZ analysis (based on coefficient of variation), do not allow identification of the consumption process dynamics and, therefore, cannot be used for the classification and improvement of forecasting models for stock consumption. This paper surveys special cases of inventory management with low demand. The results of one- and two-dimensional stock classifications are presented. The limitations of the economic order quantity (EOQ) model for inventory management strategies are determined. Methods of inventory parameter calculations for products with low demand are suggested. Integrated time series forecasting models, along with algorithms to estimate the inven- tory forecasting parameters, are proposed with regard to products with low demand. The basis for the suggested models is the following concept: all the available sources of quantitative and qualitative information should be used for managerial decision-making under uncertainty and risk. Calculations for time series with low demand are conducted for testing purposes. The obtained results confirm the adequateness of the suggested concept, aimed at solving the problem of cost reduction in supply chains.
It has been proved by the latest research on key performance indicators (KPIs) of transportation services that their successful implementation into practice is possible only if there is a thorough database of indicators and the methodology of their calculation. To reach these goals, it is necessary to classify the indicators within the framework of the system which includes the two levels: the basic (the first) and the specific (the second) KPI. This division allows to form the complex of models to calculate the basic indicators, which characterize performance (e.g. performance per hour), time parameters, expenses, reliability, etc. The article provides the analysis of papers on the methods of transportation efficiency rating in supply chains and the ways of their development to increase the efficiency of transportation; the new approach to obtain analytic dependencies to calculate KPI of transportation on the basis of the integral (factorial) method of economic analysis; the examples of calculations of some KPIs of transportation. The suggested KPI models can be used to create programs aimed at the digitalization of transportation operations in supply chains.
The article covers issues of supply chain modeling being an important step in the decision- making process. Logistics and supply chain management consider movement and transition of material flow as well as financial, information and other flows associated with it. The characteristics of a flow must be measured considering the dynamics of the flow movement. This determines the importance of simulation modeling in decision-making support system as the approach involves the implementation of modeling system algorithm functioning in a virtual time environment. The literature analysis, on the one hand, allowed to conclude that the traditional approach to functioning process characteristics determination involving consistent problem solving on the level of supply chain element, is limited or not reflecting the specifics of real processes where the parameters variability in time is possible. On the other hand, there is lack of specific recommendations on building models based on principles of system dynamics as a simulation modeling tool which allows to consider process characteristics variability. This determines the aim of the research a part of which presents supply chain simulation model illustrating the possibilities of the approach. The results of the work can be used both in practice for industrial enterprises and for future research.
Rare demand is an important part of the inventory management, nevertheless there are no any appropriate analytical descriptions or numerical examples of it except separate papers where considered the possibility to describe rare demand with Poisson distribution. The divergence between forecasts and actual data could be explained by the following reasons: the first one—extreme values in preforecasting period, as well as not significant ‘length’ of the analyzed time period; the second one—taking data for the forecast from the period of conduction certain actions (sales, promo, etc.), consideration of these actions might be done with combined forecasting methods. The paper describes the approach to assessment of inventory consumption for rare demand based on Poisson distribution. Besides, the paper contains the numerical examples and analysis of the results.
To date, a diverse array of expert assessments for quantitative (tangible) and qualitative (intangible) objects based on the analytic hierarchy process (AHP) has been accumulated in various fields of knowledge. The systematization and analysis of the collected data made it possible to put forward the hypothesis that some of the indicators (the eigenvalue of the matrix, the consistency index, and the ration of consistency) can be considered in the form of aggregates of random variables representing an intellectual product and reflecting features of human thinking. The systematization and statistical processing of the results obtained by experts on the basis of the AHP showed that the distribution functions of the values of expert estimates for the consistency indices significantly differ from the similar functions of the generated matrices.
The carried out studies show that from the point of view of the issue of efficiency increase of logistical systems, there are several key aspects. Firstly, choice of methods for managing the triad of logistics functions ‘inventory management - warehousing – transportation’, where the inventory management issues are considered as the most relevant ones. Secondly, there is recognized the need to move the studies of multi-level systems within the framework of the concept of supply chain management. Nowadays, supply chains, which are represented by the distribution system, are widespread in practice. The most common of them are two-level ones with a central supplier at the second level and a certain number of companies at the first level; and multi-level systems of the distribution configuration network in which multi-nomenclature stocks are located. The article is devoted to the design and enhancement of analytical platform for inventory management in such distribution systems.
Transportation is a key logistics function, which determines the dynamic nature of material flows in logistics systems. At the same time, transportation is a source of uncertainty of logistics operations performance in the supply chain.Obviously, the development of a new approach for evaluation of the duration of delivery “Just-In-Time” (JIT) will improve the efficiency of supply chains in accordance with one of the major criteria, namely customer satisfaction. One of the basic approaches to make effective management decisions in transportation and other logistic operations is the JIT concept. In the majority of examined sources the JIT concept is described on the verbal level without any usage of calculation dependences. The paper is devoted to the formation of analytical and simulation models, which allow obtaining the probabilistic evaluation of the implementation of unimodal and multimodal international transportation JIT. The first model where the order of the operations implementation does not affect final result is formed on the basis of the probability theory: distribution laws composition, theorems of numerical characteristics of random variables, formula of complete probability. The second model accounts the impact of operations implementation order in transportation and their interconnection and is based on the simulation (the method of statistic experiments) and shown as a corresponding algorithm, which allows to consider different limitations (technical, organizational and so on). Considered analytical dependences give the possibility to obtain the necessary estimations of the transport operations implementation according to JIT: mean transportation time, delivery implementation probability by the set moment or the delivery time with the set probability. To carry out some comparative calculations and clarify the algorithm, two international routes have been chosen: the first one is a unimodal road transportation, the second one is a multimodal transportation (road and marine transport). All the data, which is necessary for calculation has been collected on the basis of official information (in particular, the data of tachograph, special questionnaires filled in by the drivers, the survey results of the managers). For unimodal transportations analytical dependences and modelling results give close results. For the combined multimodal transportations taking into account various limitations the preference must be given to the simulation. The modelled indexes take into consideration their intercommunication and definitely estimate the supply chains reliability, and this allows decreasing the uncertainty of the logistic system.
The analysis of perspective directions of development of logistical integration allows to establish the importance of the
formation of mechanisms of management of integrated logistic functions, and also functional complexes. The
complexity of the problem is determined by the variability of the types of interaction between levels in the investigated
multi-level distributed systems, which in turn result in the variety of models of inventory management. They can be
divided into three main subgroups: the first is with independent processes, the second is with coordination and the
third – integrated models.
The most expedient way, in our opinion, to generalize the numerous AHP studies is by the formatting of a morphological
table and the corresponding block diagram. On the basis of the table, it is possible to take into account unobvious
variants that can be missed with a simple search. Thus, the AHP, despite its obvious benefits, requires further
research so as it can be successfully applied in the management of supply chains.