International Journal of Engineering Advanced Research eISSN: 2710-7167 [Vol. 4 No. 2 June 2022]
Journal website: http://myjms.mohe.gov.my/index.php/ijear
CONSOLIDATED LAST-MILE DELIVERY OPERATION TO MINIMIZE DISTRIBUTION COST
Putri Ranna1, Sutanto2 and Nahry3
1 2 3 Faculty of Engineering, University of Indonesia, DKI Jakarta, INDONESIA
*Corresponding author: [email protected]; [email protected]; [email protected]
Article Information:
Article history:
Received date : 4 June 2022 Revised date : 6 June 2022 Accepted date : 25 June 2022 Published date : 30 June 2022
To cite this document:
Ranna, P., Sutanto, S., & Nahry, N.
(2022). CONSOLIDATED LAST-MILE DELIVERY OPERATION TO
MINIMIZE DISTRIBUTION COST.
International Journal of Engineering Advanced Research, 4(2), 10-23.
Abstract: The growth of the e-commerce business increased the number of deliveries of goods to customers' homes. The problem of increasing the volume of goods to be delivered to customers' homes is the inefficient distribution system of last-mile delivery causes an increase in delivery vehicles operating in urban areas, which can add negative externalities such as congestion, noise and pollution. Another problem that often arises from last-mile delivery activities is a failure of delivery, where which will increase costs because it needs to be re-delivered to the customer's house. The implementation of a consolidation strategy in the last-mile delivery distribution system is an efficient solution to minimize distribution costs.
Consolidation did by adding a hub to the distribution system. The purpose of this study is to analyze the efficiency of the last-mile delivery distribution system with a consolidation strategy to minimize the total distribution costs by considering internal cost, external cost and opportunity costs due to the addition of a hub in the last- mile distribution system. Distribution system planning with several scenarios solved by vehicle routing problems considering different vehicle modes and time windows are known as HFVRPTW (Heterogeneous Fleet Vehicle Routing Problem with Time Windows). The VRP problem in this study was solved using VRP Spreadsheet Solver software. In addition, electric vehicles are also considered to be used as delivery vehicles is expected to reduce pollution and improve environmental quality. The result of the study succeeded in reducing distribution costs in terms of the decrease in total distance travelled and total distribution costs.
1. Introduction
The e-commerce business is predicted to grow in Indonesia over the next few years. The e- commerce business development has also boosted the growth of operators providing last-mile delivery services. Last-mile distribution in e-commerce businesses usually uses the home delivery (HD) method, and packages are sent directly to the customer/end-user location (A.A.N. Perwira et al., 2020). Last-mile delivery performance is crucial in the relationship between sellers and buyers in the e-commerce business, which directly affects the buyer's decision to reorder (Bopage, 2019).
The challenge of logistics operators providing last-mile delivery services today is to accelerate the delivery of goods for consumers while reducing the cost of delivery itself (Camalia, 2020). On the other hand, last-mile delivery is considered the most expensive, least efficient, and most polluting part of the entire logistics chain. Last-mile delivery costs can reach 53% of the total shipping costs (Dolan, 2021).
The problem with increasing the volume of parcels to be delivered to customers' homes is the inefficient distribution system for last-mile delivery. It is causing an increase in delivery vehicles operating in urban areas, which can add negative externalities, namely congestion, noise, and pollution, which negatively impact health. In addition, problems that often arise from last-mile delivery activities are failures at the time of delivery to the customer's house/end-user, where this will increase costs because it needs to be re-delivered to the customer's home.
Implementing a consolidation strategy in the last mile delivery distribution system is an efficient solution to reduce shipping costs (Simoni, 2015). In addition to consolidation, the efficiency of the distribution system to reduce transportation costs can be done by determining distribution routes to minimize the total distance traveled to optimize the use of vehicle capacity and the number of vehicles (Supardi et al., 2020). Vehicle routing problems are commonly known as Vehicle Routing Problems (VRP). VRP is defined as a problem where an optimum route plan is desired to obtain the minimum cost and travel time.
In line with the problems in last-mile delivery operations, this paper aims to analyze the efficiency of the last mile delivery distribution system with a consolidation strategy to minimize the total distribution costs by considering internal cost, external costs, possible costs due to the addition of a hub in the last mile distribution system and optimization of vehicle routes taking into account different vehicle options and time windows. This paper is organized in the following order: Section 2 explain the literature review, Section 3 is a methodology that includes data sources and processsing data, Section 4 present the results of scenario and scenario evaluation. Section 5 present conclusion and further research.
Keywords: Transportation, Consolidation, Last-Mile Delivery, Distribution System, Vehicle Routing Problem, VRP Spreadsheet Solver, Distribution Cost.
2. Literature Review
A distribution system is a method/system used to distribute goods from producers to consumers.
By setting up the right distribution system, companies get benefits such as low costs in the distribution process to high response to consumer demand (Kristina et al., 2020). Several innovations in efficiency solutions for last-mile delivery activities with the implementation of a consolidation strategy have been widely discussed by academics. In recent years, self-collection delivery has emerged as an alternative to home delivery. Based on references from research conducted by Yuen et al. (2018), “self-collection delivery involves the provision of a service point network where the operator collects and sends the recipient's package, and the consignee pays, collects, or returns the package (Piplani and Saraswat, 2012). Self-collection delivery services enable consolidated shipments, which reduce the number of road trips generated to serve customers. This reduces road congestion, demand for roadside parking, and greenhouse gas emissions and increases urban livability (Chen et al., 2017; Van Duin et al., 2016). Service points can be stationary (e.g. pick-up at locker points or convenience stores), mobile (e.g. pick-up in a vehicle equipped with a locker), attended (e.g. pick-up is assisted by service personnel), or unattended (e.g. pick-up is arranged by a fully automated system) (McKinnon and Tallam, 2003).”
In reference to research conducted by Edwards et al. (2010), “known customer pick-up (CP), allowing customers to pick up their packages from a shared delivery facility located near them conveniently, has become popular because it provides benefits on shipping costs and reduced emissions carbon (Punakivi, Yrjölä, and Holmström 2002)”. This shared delivery facility can be in the form of a parcel locker, where this innovation can reduce the number of driver stops by consolidating shipments to a single destination. In addition, the research conducted by Wang et al.
(2014) explores the competitiveness of attended home delivery, reception boxes, and pick-up points on last-mile delivery, showing the results that pick-up points are suitable for areas with a high population density of large orders. In this case, the pick-up point applied can also be in the form of a parcel locker or local depot. Parcel lockers that are provided can be self-service lockers that can allow customers to pick up packages at flexible times. The location of the parcel locker can be placed in an area that is easily accessible by customers. The application of this parcel locker can also overcome the problem of delivery failure due to the absence of customers during home delivery.
In Bosona's research (2020), the development of innovation in last-mile distribution carried out by several researchers is used to develop a last-mile delivery distribution typology. Figure 2.1 is a simplified supply chain structure.
Figure 2.1: Simplified Supply Chain Structure
S= suppliers; DC= distribution center; LDC= local distribution center; PP= pick-up point; C = customer / end user.
The red dotted line indicates the LMD focus area.
The red dotted line is the focus area for last-mile delivery. In this case, last-mile delivery, which is generally done from the distribution center directly to the customer, adds a local distribution center (LDC) and a pick-up point (PP) before going to the customer/end-user. In this case, the local distribution center (LDC) can represent a local depot, retail store, consolidation center, mobile depot, or transshipment point, depending on the different characteristics of the LMD service provider company. Pick-up points (PP) can be 'Bento box,' reception points, parcel lockers (automatic lockers that can serve 24/7), or service points such as small shops, gas stations, and train stations.
Based on the concepts and definitions explained, Bosona (2020) has developed a potential distribution model configuration for last-mile delivery, which can be seen in Table 2.1. There are three main configurations in the development of the last mile delivery distribution typology. Type I is DC-based delivery (DC delivery); Type II is LDC-based delivery (LDC delivery). Type III is PP-based delivery (PP delivery).
Table 2.1: Typology of Last-Mile Delivery Distribution System Configuration
Type Ilustration Description
Type-I (Option-1)
DC Delivery (Option-1):
Distribution from the distribution center (DC) directly to the customer's home (C) or pick-up point (PP).
Type-I (Option-2)
DC Delivery (Option-2):
Distribution from the distribution center (DC) only to the pick-up point (PP) where the customer (C) picks up their package.
Type-II (Option-1)
LDC Delivery (Option-1):
Distribution from the distribution center (DC) to the local distribution center (LDC) where the customer (C) picks up their packages.
Type-II (Option-2)
LDC Delivery (Option-2): The package is brought from the distribution center (DC) and then distributed to the customer's home (C) or pick-up point (PP).
Type III
PP Delivery: Packages are carried from the local distribution center (LDC) to the nearest pick-up point (PP) and then picked up by the customer (C).
Although applying a self-collection solution to the LMD distribution system can overcome the problem of delivery failure, several things need to be considered. Based on a reference research conducted by Tom Cherrett and Julian Allen (2019), both the mode of transportation and the nature of the journey that consumers choose when picking up their packages from their pick-up point locations will have an important influence on traffic, environmental and safety impacts associated with e-commerce last mile delivery.
2.1 Problem Statement
The efficiency of last-mile delivery is receiving increasing attention in the academic community, and more publications have appeared in recent years. Based on the literature study that has been carried out, there are gaps found in previous studies:
1. The last-mile delivery distribution system has not yet been analyzed with a consolidation strategy using several scenarios with vehicle routing characteristics using different vehicle combinations and the application of delivery time windows to minimize distribution costs.
2. Distribution costs for last-mile delivery activities by considering possible costs due to the addition of a hub in the distribution system and external costs due to negative externalities to minimize distribution costs have not been carried out.
Based on this gap, this study will focus on how much the reduction in distribution costs with optimal vehicle routing in the last mile delivery distribution system with a consolidation strategy using several scenarios with different vehicle combinations, implementing delivery time windows, and considering the possible costs due to the addition of a hub in the distribution system and external costs due to negative externalities. Vehicle routing in the last mile delivery distribution system is translated into a mathematical model and completed with the help of software.
3. Method
This study analyzes the routing of delivery vehicles in the last mile delivery distribution system with a consolidation strategy to reduce distribution costs. The research stage begins with identifying the problem and then determining the formulation of the problem and the objectives to be achieved in the research. The next step is conducting a literature study as a reference and basic knowledge related to the issues discussed in the study. The next step is to determine the research variables based on the literature study to identify the data requirements needed for research. The variables in this study consisted of independent variables and dependent variables.
The independent variables in this study are the LMD distribution system and the mode of delivery transportation. The dependent variable in this study is the total cost of last-mile delivery. The data used in this study is data on the delivery of goods from one of Indonesia's parcel delivery service companies. The data obtained contains information on the location of the depot and customers (lattitude & longitude), the number of delivery vehicles, the type of delivery vehicle, the capacity of the delivery vehicle, the delivery time window, and the delivery service time.
After all the data has been collected, the next step is to interpret and process the data for use in the next stage. The data used for analysis is delivery data at one of the Distribution Centers (DC) in the DKI Jakarta area. The data consists of seven days, starting from Monday to Sunday. Figure 3.1 describes the number of deliveries and the number of delivery vehicles used. The total delivery data for seven days is 1546 delivery points. The peak day is on Friday and off-peak day is on Sunday. The delivery vehicle used is a motorcycle (SM). The delivery time window is from 07:00 to 22:00, and the average delivery service time is 1 minute.
Figure 3.1: Total Delivery & Number of Vehicles
The next stage is to determine the scenario planning for analyzing and developing mathematical models. Scenario planning uses a variety of vehicle types and a last-mile distribution system. The distribution system refers to the Bosona (2020) literature study, described in section 2, table 2.1.
The Local Distribution Center (LDC) in the last mile distribution system is a local depot, which is a rented building, and the pick-up point (PP) is a parcel locker. The types of vehicles used are motorcycles (SM), blind vans (BV), and electric blind vans (EV). The maximum capacity for motorcycles is ten packages, and the maximum capacity for blind vans and electric blind vans is
Mon Tue Wed Thu Fri Sat Sun Total Delivery 145 280 150 268 350 226 127 Number of Vehicle 63 78 59 81 84 71 44
0 50 100 150 200 250 300 350 400
60 packages. An explanation of the development of mathematical models can be seen in section 3.1.
The next stage is applying the mathematical model using the VRP Spreadsheet Solver software from (Endorgan, 2017). VRP Spreadsheet Solver is an open-source software for vehicle routing problems. The VRP Spreadsheet Solver is operated on Microsoft Excel using the Visual Basic for Application (VBA) programming language and using the public Geographical Information System (GIS). The algorithm used in VRP Spreadsheet Solver is a metaheuristic algorithm that is Large Neighbourhood Search (LNS). The output of the VRP Spreadsheet Solver application is vehicle kilometer travelled (VKT). The internal and opportunity cost calculation refers to Mangiaracina et al. (2019), which is explained by equation 3.1.
(𝑇𝑟𝑎𝑛𝑠𝑝𝑜𝑟𝑡 𝑚𝑒𝑎𝑛 𝑐𝑜𝑠𝑡 [Rp/𝑑𝑒𝑙𝑖𝑣𝑒𝑟𝑦] (1) + 𝐷𝑟𝑖𝑣𝑒𝑟 𝑐𝑜𝑠𝑡 [Rp/𝑑𝑒𝑙𝑖𝑣𝑒𝑟𝑦] (2)
𝑃𝑎𝑟𝑐𝑒𝑙𝑠 𝑑𝑒𝑙𝑖𝑣𝑒𝑟𝑒𝑑 𝑖𝑛 𝑎 𝑑𝑒𝑙𝑖𝑣𝑒𝑟𝑦 𝑡𝑜𝑢𝑟 [𝑝𝑎𝑟𝑐𝑒𝑙𝑠/𝑑𝑒𝑙𝑖𝑣𝑒𝑟𝑦] 𝑥 (1 + % 𝐹𝑎𝑖𝑙𝑒𝑑 𝑑𝑒𝑙𝑖𝑣𝑒𝑟𝑖𝑒𝑠 [%]) +𝑂𝑝𝑝𝑜𝑟𝑡𝑢𝑛𝑖𝑡𝑦 𝑐𝑜𝑠𝑡 [ 𝑅𝑝
p𝑎𝑟𝑐𝑒𝑙] (3) (3.1)
1. Transport Mean Cost. These are the costs associated with the use of means of transport.
Transport means travel cost [Rp/km] x Travelled distance [km/delivery].
2. Driver Cost. This includes the cost of workers deliver the packages to consumers. The main components are the hourly cost of the driver and the time required (delivery time, problem- solving time, travel time).
3. Opportunity Cost. In this study, opportunity cost is a cost because it adds a hub (local depot
& pick-up point) to the last-mile distribution system. The opportunity cost is obtained by calculating the operational costs in one year divided by the average number of packages delivered in one year, then the cost is obtained in rupiah per parcel (Rp/ parcel).
The calculation for external cost refered to Cardenas (2017) which is explained by eqution 3.2 down below.
E = 𝑒𝑐𝑚 . VKTi
𝑛i (3.2)
Based on equation 3.2, E is average cost of each vehicle per parcel; ecm is a cost index; VKTi is the length of each vehicle travelled (km); ni is the number of stops per delivery. The Cost index for internal cost index is obtained by the calculation in 2022. The external cost index is obtained from Handbook on The External Cost of Transport (2019) because Indonesia has no external cost index standard. The following are the assumptions of the cost index used in the study.
Table 3.1: Cost Index
The next stage is the research analysis by comparing the total cost of the existing distribution with scenario planning and evaluating the optimal scenario in the last mile distribution system based on objective function. The last stage is making conclusions from the results of the analysis that has been carried out and suggestions for further research development.
3.1 Mathematical Model
A mathematical model is developed to form the objective function and constraints of the defined objective function. The objective function of the mathematical model that will be developed is to minimize the cost of last-mile delivery by minimizing the distance traveled, which meets all constraints. The mathematical model developed in this study considers different types of vehicles and the time window or commonly known as the Heterogenous Fleet Vehicle Routing Problem with Time Windows (HFVRPTW). The formulation of the HFVRPTW mathematical model begins by defining an objective function and then forming a constraint. Assume N = {0, 1, 2, …, n, n+1} set as customers and depot (0 and n+1 is depot notation) and ∀ij ∈N. Origin destination of trip is i and destination of trip is j. k is a vehicle of one type of vehicle m, where k = {0,1,2,3,…, Km} and k ∈ Km. Km is set of vehicles for each type of vehicle. Type of vehicle m = {0,1,2,3,…, M} and m ∈ M where M is a set of diffferent types of vehicles. The following is the formulation of HFVRPTW mathematical model.
Objective Function:
Min Z=
∑ ∑ ∑ ∑ ( 𝜶𝒎 . 𝒅𝒊𝒋𝒌𝒎 . 𝒙𝒊𝒋𝒌𝒎 + 𝜷𝒎 . 𝒅𝒊𝒋𝒌𝒎 . 𝒙𝒊𝒋𝒌𝒎 ) + (𝑶𝑪𝒎 .
𝒏+𝟏
𝒋=𝟏
𝒙𝒊𝒋𝒌𝒎)
𝒏
𝒊=𝟎 𝑲𝒎
𝒌=𝟏 𝒌∈𝑲𝒎 𝑴
𝒎=𝟏 𝒎∈𝑴
(3.3)
Constraints:
∑ ∑ ∑ ∑ 𝑿𝒊𝒋𝒌𝒎= 𝟏
𝒏+𝟏
𝒋=𝟎 𝒏
𝒊=𝟎 𝑲𝒎
𝒌=𝟏 𝒌∈𝑲𝒎 𝑴
𝒎=𝟏 𝒎𝝐𝑴
; ∀𝒊𝒋 ∈ 𝑵 ; 𝑵 ≠ 𝟎
(3.4)
∑ 𝑿𝒐𝒋𝒌𝒎 ≤ 𝟏
𝒏+𝟏
𝒋=𝟏
; ∀𝒌 ∈ 𝑲𝒎 ; ∀𝒎 ∈ 𝑴 (3.5)
∑ 𝑿𝒊(𝒏+𝟏)𝒌𝒎 ≤ 𝟏
𝒏
𝒊=𝟎
; ∀𝒌 ∈ 𝑲𝒎 ; ∀𝒎 ∈ 𝑴 (3.6)
Internal cost Index External Cost Index
(Rp/km) (Rp/km)
SM Rp2.324 Rp304
BV Rp5.123 Rp526
EV Rp6.777 Rp0
Type of Vehicle
∑ 𝑿𝒊𝒌𝒎𝒈 −
𝒏
𝒊=𝟎
∑ 𝑿𝒋𝒌𝒎𝒈= 𝟎
𝒏+𝟏
𝒋=𝟏
; ∀𝒈 ∈ 𝑵 (3.7)
∑ ∑ 𝒅𝒋. 𝑿𝒊𝒋𝒌𝒎 𝒏+𝟏
𝒋=𝟏
≤ 𝑸𝒎 ; ∀𝒌 ∈ 𝑲𝒎 ; ∀𝒎 ∈ 𝑴
𝒏
𝒊=𝟎
(3.8)
𝒕𝒔 ≤ 𝒕𝒐𝒌 (3.9)
𝒕𝒐𝒌′ ≤ 𝒕𝒆 (3.10)
𝒕𝒐𝒌′ = 𝒕𝒐𝒌+ ∑ ∑{𝒕𝒊𝒋𝒌𝒎+ 𝒕𝒄𝒋} ;
𝒏+𝟏
𝒋=𝟏 𝒏
𝒊=𝟎
∀𝒌 ∈ 𝑲𝒎 ; ∀𝒎 ∈ 𝑴 (3.11)
𝒙𝒊𝒋𝒌 ∈ {𝟎, 𝟏} ; ∀𝒊𝒋 ∈ 𝑵 ; ∀𝒌 ∈ 𝑲𝒎 ; ∀𝒎 ∈ 𝑴 (3.12)
The Objective function of HFVRPTW is to minimize the total cost of Z which includes internal cost, external cost and opportunity cost. Internal cost is obtained by multiplying the internal cost index α𝑚 with the total trip length dijkm. For the external cost is obtained by multiplying external cost index β𝑚with the total trip length dijkm. The opportunity cost is a cost because it adds a hub (local depot & pick-up point) based on last-mile distribution scenarios. Total length of trip is the sum of trip length of each vehicle k, vehicle type m. Constraints (3.4) show that each customer is visited only once. Constraints (3.5) show that all vehicles start from the depot and constraints (3.6) show that all vehicles end to the depot. Constraints (3.7) is the balance of vehicle current where the amount of current in the depot minus the amount of current out of the depot must be equal to 0, g is the current of the vehicle. Constraints (3.8) is the carrying capacity of a vehicle k, dj is the number of customer packages j and Qm is the maximum number of packages of the types of vehicle k. Constraints (3.9), (3.10) and (3.11) are the time window constraints where ts is the earliest time to start operating delivery vehicle; te is the latest time to start operating delivery vehicle; tcj is the loading time and waiting time for customer j; tok is the earliest departure time from the depot; t’ok
is the latest arrival time to the depot; tijkm is the travel time of the vehicle k from the location i to the location j. Constraint (12) is a binaray variable that only has value 0 or 1; if Xijkm=1 then route ij is passed by vehicle k of vehicle type m.
4. Results and Discussion
The mathematical model of HFVRPTW was solved using VRP Spreadsheet Solver in the existing conditions and scenarios. The scenario planning on the last-mile delivery distribution system considers the independent variables, namely the type of vehicle and the last mile distribution system. The delivery vehicles considered are motorcycles (SM) and blind vans (BV). The distribution system considered refers to the Bosona reference (2020) described in section 2, table 2.1, namely Type 1 DC delivery (Option 1 & Option 2), Type 2 LDC delivery (Option 1 & Option 2), and Type 3 PP delivery. Scenario planning represent in Table 4.1.
Table 4.1: Scenario
The location of the hub (LDC & PP) in the last-mile distribution system is determined using a zoning approach, represent in Figure 4.1.
Figure 4.1: Location of LDC and PP Skenario Jenis Kendaraan
Pengiriman Sistem Distribusi
Eksisting SM DC delivery
SC1 SM DC delivery (Opsi 1)
SC2 SM, BV DC delivery (Opsi 1)
SC3 SM DC delivery (Opsi 2)
SC4 BV DC delivery (Opsi 2)
SC5 SM, BV DC delivery (Opsi 2)
SC6 SM LDC delivery (Opsi 1)
SC7 BV LDC delivery (Opsi 1)
SC8 SM, BV LDC delivery (Opsi 1) SC9 SM, BV LDC delivery (Opsi 2)
SC10 SM, BV PP delivery
PP LDC
The application of the mathematical model using the VRP Spreadsheet Solver software in the existing conditions and scenarios results in the value of vehicle kilometers traveled (VKT), as shown in Figure 4.2.
Figure 4.2: Vehicle Kilometre Travelled
Based on Figure 5, the value of Vehicle Kilometre Travelled (VKT) in the existing condition is much higher than in the ten scenarios. The high value of VKT in existing conditions is due to the lack of efficiency by logistics operators. This VKT value will be used to calculate internal cost and external costs based on the cost index data in Table 3.1. For opportunity costs, the calculation results show that the additional cost is IDR 2,517/parcel due to adding a parcel locker hub to the last-mile distribution system. Meanwhile, the addition of a local depot hub is Rp. 886/parcel. Total distribution costs can be calculated by adding up internal, external, and opportunity costs. The total distribution costs during peak day and off-peak day are shown in Figure 4.3.
Figure 4.3: Total Distribution Cost
Eksisting SK-1 SK-2 SK-3 SK-4 SK-5 SK-6 SK-7 SK-8 SK-9 SK-10
Peak Day V1 1866.30 296.89 42.44 324.36 0 20.9 323.09 0 65.69 250.26 227.07
Peak Day V2 0 0 97.35 0 106.33 94.71 0 96.16 66.39 96.16 96.16
Off Peak Day V1 662.67 124.94 16.19 129.86 0 70.61 140.93 0 11.81 133.76 105.98
Off Peak Day V2 0 0 61.92 0 68.19 27.17 0 47.13 31.62 47.13 47.13
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Vehicle Kilometer Travelled (VKT)
Peak Day V1 Peak Day V2 Off Peak Day V1 Off Peak Day V2
Rp- Rp1,000,000 Rp2,000,000 Rp3,000,000 Rp4,000,000 Rp5,000,000 Rp6,000,000
Total Cost
Scenario
Peak Day Off Peak Day
Based on Figure 4.3, the lowest total distribution cost compared to other scenarios is the SC-7 scenario. Compared to the existing conditions, the total distribution costs were reduced by 83.14%
on peak days and 78.25% on off-peak days. This happens because, in the SC-7 scenario, deliveries are carried out using blind vans only to 4 LDC (local depot) locations, resulting in a small VKT value, whereas in this scenario, shipments are consolidated, and delivery points are reduced.
Table 4.2: Percentage Reduction in Total Costs Compared to Existing
To reduce the external impact and improve the environment, using electric blind vans in last-mile delivery is promising to use as delivery vehicles. In this study, the use of electric blind vans as delivery vehicles reduced the total distribution costs by 81.5% on peak days and 79.45% on off- peak days compared to the existing conditions in the SC-7 scenario. In general, SC-7 is the best scenario with the most considerable percentage reduction in total distribution costs. The distribution system on the SC-7 is distribution carried out from the distribution center (DC) to the local distribution center (LDC), where the customer (C) can pick up their packages. The consolidation of the last mile distribution system in this scenario has succeeded minimize the total distribution costs and vehicle travel distances.
5. Conclusion
Based on the analysis and discussion of the research, the last-mile delivery distribution system with a consolidated strategy on HFVRPTW using a case study at a parcel delivery company in Jakarta, the results show that SC-7 can reduce total distribution costs by 83.14% on peak days and 78.25% on off-peak days and reduce vehicle travel distances compared to the existing conditions.
Distribution carried out from the distribution center (DC) to the local distribution center (LDC), where the customer (C) can pick up their packages. The use of electric vehicles can be considered to reduce the impact of negative externalities due to delivery activities. Distribution system of last- mile using consolidation strategy is efficient to minimize the total distribution costs and vehicle travel distances.
peak day off-peak day
Total Total
1 Eksisting 0.00% 0.00%
2 SC-1 70.24% 66.69%
3 SC-2 72.66% 63.02%
4 SC-3 66.20% 62.05%
5 SC-4 71.33% 59.53%
6 SC-5 71.55% 62.18%
7 SC-6 76.91% 72.27%
8 SC-7 83.14% 78.25%
9 SC-8 83.05% 81.50%
10 SC-9 61.66% 50.07%
11 SC-10 60.34% 50.36%
No. Scenario
Total Cost
This study has a limitation; looking ahead, the easiest way for customers to reduce their collection time and distance is to make more pick-up points it can be placed in a convenient location. For future research, optimization of distribution systems on last-mile activities can be developed using other innovative solutions that could improve shipping goods' performancevery in urban areas.
6. Acknowledgement
This research was supported by Directorate of Research and Development of University of Indonesia, 2022.
References
A.A.N Perwira Redi et al. (2020). A Simulated Annealing Algorithm for Solving Two-Echelon Vehicle Routing Problem with Locker Facilities. MDPI Algorithms.
Bopage, Gayani et al. (2019). A Strategic Model to Improve the Last Mile Delivery Performance in E-Commerce Parcel Delivery. Proceedings of the International Conference on Industrial Engineering and Operations Management Bangkok, Thailand, March 5-7, 2019.
Bosona, Techane. (2020). Urban Freight Last Mile Logistics—Challenges and Opportunities to Improve Sustainability: A Literature Review. MDPI. Sustainability 2020, 12, 8769.
Camalia, Melisa (2020). Breaking Down Last Mile Delivery: Konsep, Tren, dan Solusi. Breaking Down Last Mile Delivery: Konsep, Tren, dan Solusi - Mile Blog
Cárdenas I, Beckers J, Vanelslander T. (2017). E-commerce last-mile in Belgium: Developing an external cost delivery index. Res Transp Bus Manag, vol. 24, no. 12, 3–9
Dolan, Shelagh. (2021). The challenges of last mile delivery logistics and the tech solutions cutting costs in the final mile. Businessinsider. https://www.businessinsider.com/last-mile-delivery- shipping-explained?r=US&IR=T
Edwards, J., A. McKinnon, T. Cherrett, F. McLeod, and L. Y. Song. (2010). “Carbon Dioxide Benefits of Using Collection-Delivery Points for Failed Home Deliveries in the United Kingdom.” Transportation Research Record 2191: 136–143.
Erdogan, Gunes. (2017). An open source Spreadsheet Solver for Vehicle Routing Problems.
Computers and Operations Research, 84, 62–72.
European Commission. (2019). Handbook on The External Costs of Transport. European Commission, Brussels.
Kristina, Sonna. Et al. (2020). Penerapan Model Capacitated Vehicle Routing Problem (CVRP) menggunakan Google OR-Tools untuk Penentuan Rute Pengantaran Obat Pada Perusahaan Pedagang Besar Farmasi (PBF). Jurnal Telematika, Vol. 15 no.2. Institut Teknologi Harapan Bangsa, Bandung.
Mangiaracina, Alessandro., Arianna Seghezzi, Perego., and Tumino, Angela. (2019). Innovative solutions to increase last-mile delivery efficiency in B2C e-commerce: a literature review.
Department of Management, Economics and Industrial Engineering, Politecnico di Milano, Milano, Italy. www.emeraldinsight.com/0960-0035.html
Simoni, M. D., Bujanovic, P., Boyles, S. D., & Kutanoglu, E. (2018). Urban consolidation solutions for parcel delivery considering location, fleet and route choice. (Case Studies on Transport Policy, 2018).
Supardi, Edi., Ruben Chandra Sianturi. (2020). Metode Saving Matrix Dalam Penentuan Rute Distribusi Premium Di Depot SPBU Bandung. Jurnal Logistik Bisnis, Vol. 10, No.1, Mei 2020.
Tom Cherrett and Julian Allen. (2019). Last mile urban freight in the UK: how and why is it changing? Future of Mobility: Evidence Review. Government Office for Science.
https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_dat a/file/777682/fom_last_mile_road_freight.pdf
Yuen, et al. (2018). An investigation of customers’ intention to use self-collection services for last- mile delivery. Transport Policy.
