SIMULATION MODELING OF A DIGITAL CROWD-LOGISTICS «LAST-MILE»DELIVERY PLATFORM IN THE SHARING ECONOMY

Abstract

The article investigates the directions of digital platform transformation aimed at optimising logistics activity under the conditions of the sharing economy. The relevance of the study is determined by the rapid growth of e-commerce and the increasing share of last-mile delivery in total logistics costs, which, according to recent estimates, has risen from 41 % to 53 % of total shipping costs, as well as by the expansion of crowdshipping as a logistics manifestation of the sharing economy, in which private individuals deliver parcels through a digital platform using their own vehicles and incidental routes. The purpose of the article is to develop an agent-based simulation model of a hybrid urban courier delivery system, in which a digital platform combines dedicated couriers of a delivery service and private crowd carriers, and to quantify the effects of such a transformation. The model was built in the AnyLogic 8 Personal Learning Edition multimethod simulation environment using the agent-based approach, agent statecharts and a GIS map of the city. The behaviour of a crowd carrier is described by a statechart with the states of waiting, decision-making, moving to the pickup point and moving to the customer, while the probability of accepting an offered order is set as a model parameter. A series of simulation experiments was carried out for three delivery scenarios: a traditional one relying solely on dedicated couriers, a hybrid one with a high willingness of crowd carriers to accept orders, and a hybrid one with a low willingness. The simulation results showed that connecting crowd carriers through a digital platform reduces the average delivery time by 27.5 %, increases the share of on-time deliveries from 78.4 % to 94.1 % during peak periods and decreases the unit cost per delivery by 18.9 %, mainly due to the flexible absorption of peak loads and the replacement of fixed labour costs with variable per-delivery payments. At the same time, the experiments quantitatively confirmed the critical dependence of platform efficiency on reaching a critical mass of active carriers: with a low acceptance probability, the advantages of the hybrid model are almost eliminated. The practical value of the results lies in the possibility of using the model as a decision support tool for determining the rational ratio of dedicated and crowd carriers, testing community motivation mechanisms and assessing the consequences of demand changes without risky field experiments.