Last Mile Delivery With Stochastic Travel Times Considering Dual Services

Abstract

Nowadays there are two prevailing delivery modes in the last mile delivery; one is the home delivery (HD) that vehicles deliver parcels to customers' homes; and the other is the customers' pickup (CP) that vehicle deliver parcels to some kind of intelligent express boxes where customers can pick up their parcels with free time. This article studies a green vehicle routing problem considering dual services (HD and CP) with stochastic travel times (GVRP-DS-STT) to provide customers with sustainable and diversified delivery services. The GVRP-DS-STT problem is formulated as a two-stage stochastic optimization model with recourse strategy. The purpose of the model is to minimize the total operational cost under stochastic environment. In addition, a two-stage heuristic algorithm integrating with a sampling strategy is developed to solve approximately the problem, the first of which is the greedy-based initial feasible solution generation, and the second of which is an improvement heuristic with late acceptance to explore the solution space. The computational results show that there are increasing benefit in terms of total operational cost, number of the vehicles used and loading rate with the increase of percentage of CP customers. The time windows have great effect on the operational cost, but more CP customers can reduce their impact. The stochastic model outperforms the deterministic model in terms of total operational cost. INDEX TERMS Green vehicle routing problem, stochastic travel times, dual services, sample average approximation sampling, hybrid heuristics.

Data-driven optimization for last-mile delivery

“…Jiang et al [14] consider a similar last-mile delivery problem and aim to reduce the total costs and carbon emissions, their model is a variant of the traveling salesman problem. Zhou et al [15] study a green VRP considering dual last-mile delivery services with stochastic travel times. However, both of them do not consider the prediction problem with multivariate data.…”

“…Assuming there are p = 5 features, and the nodes-vehicles pair is (15, 3) in which further entry is a total number of customers and the latter is the number of total drivers. The customer demand is random generated as d i = [24,20,20,25,24,13,16,20,25,25,16,17,22,19,15] and vehicle capacity vector is Q = [94, 108, 100] . We generate customers' positions in a 100*100 region and suppose the initial deport is situated in the central position.…”

Service Quality Evaluation of Terminal Express Delivery Based on an Integrated SERVQUAL-AHP-TOPSIS Approach

“…It has become an indispensable part of daily life. Taking the terminal distribution in university campus as an example, almost every logistics service firm has its service dot and serves for faculties' delivery demands [8].…”

“…e innovative VRP applications and distribution modes contribute to efficiency improvement of express delivery [3,6]. e double service including home delivery and customers' pick-up are taken into consideration during the VRP optimization [7,8]. Zhou [9] developed a novel nonlinear programming model to assist logistics managers to find an optimal integrated location-routing solution by an improved GA algorithm, and the numerical case showed that it can improve the efficiency of last mile delivery by the formulated model.…”

An Effective Heuristic for Multidepot Low-Carbon Vehicle Routing Problem

“…Genetic algorithm (abbreviated form GA) has the advantage of fast convergence in solving NP-hard problems, so it has become an important method for solving vehicle routing problems [28][29][30][31]. In order to verify the effectiveness of IFFO, the paper compares the solution results of FFO, IFFO, and GA and draws the distribution route map of the three algorithms.…”