Razvoj simulacionih modela dostave pošiljaka u skladu sa konceptom izmeštanja poslovnih procesa

Abstract

Istraživanje koje je prikazano u ovoj doktorskoj disertaciji ima za cilj da popuni prazninu koja postoji u polju simulacija koje verodostojno prikazuju ponašanje sistema u kojima su dostavni procesi potpuno izmešteni (to jest kada se radi o crowdosurced dostavi pošiljaka), a koji se oslanjaju na dostavljače koji koriste heterogene vidove dostave (to jest automobil i bicikl). U tom smislu, razvijeni su simulacioni modeli koji omogućavaju merenje perfomansi sistema dostave pod različitim uslovima i analizirani su faktori koji imaju najveći uticaj na kretanje vrednosti njihovih ključnih indikatora performansi (iskazanih kroz ukupan obim realizovanih dostava i procenat dostava realizovanih u vremenski predviđenom roku). Poseban fokus je stavljen na verodostojnost simulacionih modela, tako da se u istraživanju umesto standardne aproksimacije bitnih parametara simulacije kao što je to slučaj u sličnim studijama, koriste podaci iz realnih uslova upotrebom popularnih servisa koji omogućuju planiranje ruta i navigaciju.

The research presented in this doctoral dissertation is aimed at filling the gap present in the field of the simulations that credibly show the behavior of the systems in which delivery processes are completely dislocated (i.e. when crowdsourced item delivery is concerned), which rely on the couriers who use the heterogenous forms of delivery (namely a car and a bicycle). In that sense, the simulation models that enable the measurement of delivery system performances under different conditions have been developed and the factors exerting the biggest influence on the movement of their key performance indicators (expressed through the total volume of realized deliveries and the percentage of the deliveries realized within the envisaged time window) have been analyzed. A special focus has been put on the credibility of the simulation models, so that the data obtained from real conditions using the popular services that enable route planning and navigation are used in the research study instead of the standard approximation of the important simulation parameters, as is the case in similar studies.