Modeliranje rizika u logističkim procesima sa primenom u transportnom osiguranju

Abstract

Logistički sistemi i procesi su veoma izloženi brojnim rizicima koji mogu nastati na osnovu realizacije različitih negativnih scenarija događaja. Osnovni razlog za ovu činjenicu je prisustvo velikog broja subjekata logističkih sistema, postojanje različitih interakcija između velikog broja podsistema ili podprocesa, što utiče na nastanak poremećaja i neizvesnosti, kako na lokalnom nivou, tako na nivou celokupnog sistema. Menadžment rizika je naučna disciplina koja se bavi analizom, ocenom i kontrolom rizika. Ova disciplina je sve više razvija u mnogim oblastima, uključujući logistiku. Često je nemoguće predvideti i kontrolisati sve potencijalne rizike koji su prisutni u logističkim sistemima ili procesima, ali je moguće sagledati najznačajnije i svesti ih na najmanju moguću meru. U praksi se koriste različite metode menadžmenta rizika koje se primenjuju u zavisnosti od brojnih internih i eksternih faktora i različitih ograničenja. Modeliranje rizika predstavlja dinamički proces koji podrazumeva multidisciplinarna znanja i širok spektar aktivnosti u cilju izrade primenljivog, korisnog i kredibilnog modela za analizu i ocenu rizika. Polazna ideja istraživanja je da se ukupan rizik logističkih procesa modelira na osnovu postupka vrednovanja značaja različitih elementa rizika, njihovog međusobnog odnosa i uticaja na ukupan rizik. U ovoj disertaciji su predloženi modeli analize i ocene rizika zasnovani na metodama Analitičkog Hijerarhijskog Procesa (AHP), Fazi Analitičkog Hijerarhijskog Procesa (FAHP) i fazi logike. Sve navedene metode se koriste u uslovima nedostatka reprezentativnih statističkih podataka koji su neophodni u cilju donošenja pouzdanih upravljačkih odluka. Metode podrazumevaju prikupljanje i formalizaciju ekspertskog znanja, koje se implementira u odgovarajuće modele. Veoma zastupljen instrument menadžmenta rizika u logistici je institucija transportnog osiguranja. Kako su rizici u logističkim procesima i rizici koji se pokrivaju transportnim osiguranjem u velikom delu komplementarni, ocena rizika u ovim oblastima može biti zasnovana na istim ili sličnim matematičkim i statističkim pretpostavkama, metodama ili modelima. U skladu sa postavljenim ciljem istraživanja, u disertaciji je dato tumačenje rizika kao kompleksne veličine, opisane su osnovne karakteristike menadžmenta rizika u logistici, predložen je postupak modeliranja rizika u logističkim procesima i postupak primene modela u transportnom osiguranju. Izrađena su tri modela za ocenu rizika, koji su primenljivi u logistici i osiguranju. Svi modeli su testirani na numeričkim primerima, i rezultati testiranja su pokazali ispravnost polaznih pretpostavki za njihovu izradu i primenljivost u praksi. Predloženi modeli za analizu i ocenu rizika u logističkim procesima i njihova primena u transportnom osiguranju predstavljaju novi pristup, koji može naći značajnu primenu u praksi.

Logistic systems and processes are very exposed to numerous risks that may occur due to various negative scenarios of events. The main reason for this fact is presence of numerous subjects of logistic systems, existence of various interactions between big number of sub-systems or sub-processes, which causes disturbances and uncertainties, both locally and at the level of the entire system. Risk management is a scientific discipline dealing with analysis, assessment and control of risk. This discipline is more and more developed in numerous areas, including logistics. It is often impossible to predict and control all potential risks that are present in logistic systems or processes, but it is possible to review the most important ones and reduce them to the lowest possible extent. In practice, various risk management methods are used, depending on numerous internal and external factors and various limits. Risk modelling is a dynamic process that means multidisciplinary knowledge and wide assortment of activities for the purpose of making an applicable, useful and credible model for analysis and assessment of risk. The initial idea of a research is to model total risk of logistic processes on the basis of process of evaluation of significance of different risk elements, their inter-relations and influence on total risk. This dissertation contains proposed models of analysis and assessment of risk based on Analytic Hierarchy Process models (AHP), Fuzzy Analytic Hierarchy Process (FAHP) and Fuzzy Logic. All stated methods are used in cases where representative statistical data are missing, which are necessary for making reliable management decisions. Methods shall mean collection and formalization of expert knowledge that is implemented in corresponding models. Risk management instrument that is highly present is institution of transport insurance. Risks in logistic processes and risks covered by transport insurance are greatly complementary, so assessment of risk in these areas can be based on the same or similar mathematical and statistical assumptions, methods or models. In accordance with set objective of a research, this dissertation contains interpretation of risk as a complex value, main characteristics of risk management in logistics are described, risk modeling procedure for logistic processes is proposed as well as procedure for application of models in transport insurance. Three models for risk assessment were made, which are applicable in logistics and insurance. All models are tested on numerical examples and tests results shown accuracy of initial assumptions for their making and applicability in practice. Proposed model for analysis and assessment of risk in logistic processes and their application in transport insurance present a new approach, which can have a significant application in practice.