Sustainability Aspects of Drone-Assisted Last-Mile Delivery Systems—A Discrete Event Simulation Approach

Abstract

The goal of this article is to examine the advantages and disadvantages of the application of drones in last-mile delivery systems from a sustainability point of view regarding CO₂ emissions and energy consumption. As commercial drones are developing rapidly, the application of such tools in the field of last-mile delivery and transportation can offer many opportunities to increase service flexibility, reduce delivery time and decrease CO₂ emissions and energy consumption. In this article, a discrete event simulation is applied to examine how the assistance of drones in parcel delivery services can influence the sustainability aspects of such services regarding CO₂ emissions and energy consumption in an urban environment. Based on factory parameters, a vehicle-based delivery scenario is compared to a drone-assisted scenario under ideal conditions. According to the results, within the simulation parameters with the assistance of drones, a decrease in CO₂ emissions and energy consumption is possible in last-mile delivery services, but more environmental, technological and financial limitations should also be addressed and incorporated to determine whether such a development is worthwhile from a last-mile delivery company’s point of view.

1. Introduction

Drone technology is one of the most innovative solutions today. It has a wide range of uses, from infrastructure [1] to agriculture [2,3,4], construction [5], transport [6], military and media [7,8]. In each of these areas, drones have different specifications that make them extremely versatile. They also have the potential to be very useful on the CEP (courier, express, parcel) market, providing a flexible transportation alternative in last-mile delivery services in crowded urban areas on a larger scale. The appropriate design and optimization [9] of such transportation systems requires considering several aspects in terms of logistics [10], scheduling and distribution [11] to be able to operate on a required level, and environmental sustainability is also one of those aspects.

The goal of this article is to examine the advantages and disadvantages of the application of drones in last-mile delivery systems from a sustainability point of view regarding CO₂ emissions and energy consumption to determine if drones can be beneficial in reducing emission an consumption, resulting in a more environmentally sustainable delivery service.

First, with the help of a literature review, this paper explores the development and application of modern-day drones in a variety of processes, highlighting how they can be applied in last-mile delivery transportation systems. Second, with emission and consumption parameters of a popular delivery vehicle and drone type, which have been provided by buildings and factories, this paper examines the differences between a vehicle-based and a drone-assisted last-mile delivery process by using a discrete event simulation model and comparing the simulation results of the two scenarios with each other. Finally, limitations of the current scenario simulations and future research directions are discussed and conclusions are drawn.

1.1. The importance of Drone Technology

Companies and individuals are continuously trying to exploit and explore the potential of drones [12]. Because of this, the drone market can achieve great economic results not only through services, but also through the sale of drones and accessories, which mainly includes batteries [13]. In addition to batteries, the drone trade also covers cameras, sensors and data processing software, which, after batteries, also represent a large portion of the costs associated with drones. In the drone services sector, the area of inspection and maintenance is the most exploited [14], while location services are considered the least exploited [15,16].

Forecasts that with the expected economic growth of the drone market [17], market revenues could reach USD 43 billion by 2025. While drone technology is becoming more widespread worldwide, Asia and North America are leading the market in the application of the technology. In Europe, the uptake of drones has only recently started, which means that their uptake in Hungary is also recent [18]. This is also reflected in Hungarian regulations, as the latest government regulation on drones came into force in 2021 [19]. In this document, all issues regarding the use of drones in the country are regulated. As in Hungary, the most recent regulation on drones in the European Union was adopted in 2019, which also shows that the finality of the regulations has been reached in the last few years. As a result, the growth of the drone market in Europe can be justified by all factors, as all conditions (economic, legal) have been provided for the expansion of drone services and drone trade. There have been a number of concerns about the environmental impact of drone technology [20] and the cost of batteries [21]. Our research aims to provide insight into these issues. While we have focused mainly on the negatives, it is also worth presenting some of the positives that can be applied to all areas of use. These include access to hard-to-reach places [22], mobility, high-quality data generation and work efficiency [23,24], all of which are due to the innovative nature of the technology and the fact that drones allow the use of different sensors and cameras as long as the weight parameters are compatible [25,26].

1.2. Drone Technology in Last-Mile Delivery

Drones are used in many industries today: agriculture, defense, environment, transport development, construction, as well as in the assessment of infringement cases. Drones are expected to be used more frequently in the near future to perform certain delivery services [27]. Nowadays, the last-mile delivery process is classically achieved by a truck [28], but the use of drones in transport/delivery is becoming more widespread. There are ultimately two main methods for drone delivery: in one case, the drone leaves the warehouse, delivers the package and then returns to the point of origin; in the other case, the drone is linked to a van, which also acts as a warehouse or a loading station for the drone [29]. The benefits of drone delivery include reduced congestion [30], pollution, delivery time and costs [28]. It is a non-negligible fact that consumer preferences have been transformed by the growth of e-commerce, as consumers are increasingly expecting faster delivery services. However, this phenomenon poses a major challenge to the courier, express and parcel (CEP) market [31]. Fast and on-time delivery is also associated with increased consumer satisfaction [32].

More companies are using drones for urban deliveries, such as Amazon Prime Air, or the classic parcel delivery companies FedEx, DHL, Wing and UPS. These companies are using drones for last-mile small parcel deliveries (max 5 kg). The reason why the use of drones in last-mile delivery is advantageous is that the efficiency of parcel delivery in terms of time and cost efficiency can be increased most easily with drones. This final process can account for up to 28% of the total transport cost and therefore has a relatively high cost yield [33]. Using drones in last-mile delivery can help reduce delivery time, improve efficiency and reduce carbon emissions [34]. An extensive and thorough literature review of drone-based parcel delivery logistics systems and their feasibility was conducted [30]. The authors identified several challenges associated with drone-based delivery systems, including technical limitations, regulatory barriers and public acceptance. The study concludes by highlighting the need for further research and development to overcome these challenges and to realize the full potential of drone-based delivery systems. Thus, it is undeniable that further research is needed on the topic of last-mile delivery with drones to determine whether this method is truly better preferred over traditional truck delivery. Sustainability and environmental considerations should not be left out of the analysis of last-mile delivery by drones [35,36]. It can be concluded that, in cases where the warehouse is close to the final delivery point, drone delivery has lower CO₂ emissions than truck delivery along the same route [36]. A similar conclusion was reached in the study by Elsayed and Mohamed (2020), which confirms the relevance of last-mile delivery by drone compared to ground delivery methods [37]. According to their results, drone delivery has better CO₂ emissions than the average delivery by a fuel driven car or even an electric car [38], which is in line with the European Commission’s 2019 [39] commitment of 0.147 kg CO₂ emissions per kilometer per order. However, it has been proven that delivery by drone is not always preferable over delivery by truck. An important aspect to consider when planning the route is to take into account the different nodes; the waypoints or rendezvous locations should be predefined for both drones and trucks [40].

In the present study, we address a similar question of whether drones can be beneficial in last-mile delivery, first and foremost from a sustainability perspective. Since a simulation model is used in the study, we consider it important to present publications with similar methodologies. An interesting result was obtained in a study that also presented a freight operations modelling approach for urban delivery and pickup with flexible routing. The approach incorporates discrete event simulation and GIS to enable the real-time tracking of freight vehicles and dynamic rerouting of delivery routes [41]. The study found that the approach can improve the efficiency of urban freight operations by reducing delivery time and increasing delivery frequency. The authors also identified several challenges associated with the implementation of the approach, including data availability, computational complexity and user acceptance. The study concluded by suggesting further research to address these challenges and improve the practicality of the approach. In contrast to the previous method, a sequential optimization–simulation approach for planning the transition to a low-carbon freight system was applied in a New Zealand study [42]. The approach involves identifying the most efficient and sustainable transport modes for different types of freight and developing a phased transition plan to gradually replace high-emission transport modes with low-emission alternatives. The study applied the approach to a case study on the North Island of New Zealand and found that it can help reduce emissions and improve the efficiency of freight operations. The authors also identified several challenges associated with the approach, including the complexity of the transport system and the need for stakeholder engagement. The study concluded by recommending further research to refine the approach and improve its practicality for real-world implementation. In addition to the use of simulation models, we also find research using mathematical models, such as solving the truck and drone cooperation problem, for which a mathematical model was introduced using a simulated annealing algorithm [40]. In this simulation, the truck assists the drone in replacing and recharging the battery. However, in addition to truck-based charging, other options could be considered, such as touch-based charging stations on the roofs of buildings [43]. Demonstrated the effectiveness of their proposed solution in Simulink software using a simulation [44]. To extend the limited battery life of drones, charging stations placed on buildings help to achieve longer runtimes.

In summary, there are several studies on last-mile delivery with drones, most of which take into account sustainability aspects (CO₂ emissions) and use discrete event simulation models, such as in this study.

2. Materials and Methods

The aim of the research is to examine the advantages and disadvantages of the application of drones in last-mile delivery systems from a sustainability point of view regarding CO₂ emissions and energy consumption. As delivery with drones is becoming increasingly popular, the question arises, taking into account costs and environmental impact (CO₂, EF), whether it is truly more beneficial compared to delivery by van. Discrete event simulation is applied to examine how the assistance of drones in parcel delivery services can influence the sustainability of such services from CO₂ emission and energy consumption perspectives in an urban environment, while also highlighting the possible challenges of such drone-assisted systems for delivery companies.

Two scenarios were created and compared with each other. In the first scenario, a regular vehicle-based delivery system is considered, while in the second scenario, the last-mile delivery system is also assisted by drones.

2.1. Scenarios

In the methodology, we consider the following case: the Hungarian city of Győr is used as a basis, with a population of 124,685 people. Taking into account the cartographic shape of the city, we created our concept, assuming that there is a parcel distribution center in the center of the city, which distributes and delivers parcels of such size, which can also be delivered by the drone included in the simulation (DJI Matrice 300 RTK). Afterward, we identified 5 smaller distribution stations from where last-mile delivery can be carried out with drones. In the first scenario, only the van (Fiat Ducatio BEV 3.5t LH2 122LE) distributes the packages to the customers from the center. In the combined scenario, the van delivers the packages from the center to the smaller stations, from where the drones deliver the packages to the customers.

For the simulation, we use the energy consumption and emission data of both vehicles according to their factory specifics. In the simulation, each of the endpoints marked in last-mile delivery indicate 10 customers. Thus, we display a total of 150 deliveries in our work. In our research, we compare two scenarios with the assigned parameters and limitations:

• Scenario (A)

In Scenario (A) (Figure 1), the packages are distributed between the customers from the distribution center by van.

• Scenario (B)

In Scenario (B) (Figure 2), the packages are distributed between the drone stations by van from the distribution center, then drones provide the last-mile delivery service from the stations to the customers.

The construction of drone stations also involves costs; however, at this initial stage, we did not include this cost element in the simulation, nor did we include the costs of purchasing and maintaining the drone and the car. We will include these cost elements in subsequent more detailed research. In this study, the aim is to examine which of the two scenarios are the most energy-efficient in terms of delivery, and which scenario has the least burden on the environment.

The calculations are based on factory data and technical parameters certified by the manufacturers [45] and discrete event simulation is used to determine which scenario is the best solution in terms of cost-effectiveness, CO₂ and EF emissions [46]. Furthermore, the studies of Elsayed and Mohamed (2020) formed the basis of the logistics implementation and problems [37]. At the end of our research, we propose a new interpretation of last-mile delivery.

2.2. Discrete Event Simulation

The main feature of the discrete event simulation model (DES) is that the stated variables change only at discrete points in time, that is, they change only in a discrete set of points in time. In this respect, the DES method is the opposite of continuous simulations, since, here, the time jumps from one scheduled event to the next, rather than continuously spinning. Events integrated in the simulation can schedule additional events. Simulations are essentially simplifications of reality, focusing on the system as a whole. Simulations are most often performed using software that has been developed specifically for a particular purpose. The aim of a simulation is to represent reality as closely as possible and thus provide valuable information to the user. The correct implementation of these processes can improve the efficiency of the individual processes and perform optimization tasks [20].

In addition to striving for simplicity, simulation models should be detailed enough to mimic reality as closely as possible [47,48]. The specification of different parameters in simulation software is essential, as they allow the software to illustrate each case in the most realistic way. These applied frameworks ensure the description and operation of the models [49].

2.3. Parameters

In the present study, a discrete event simulation software was used to simulate how efficiently a truck and a drone can interact in delivery processes. Discrete event simulation also helps to find the strengths and weaknesses of the truck–drone collaboration. Table 1 contains the factory parameters of the vehicle and drone which were used in the simulation framework and the following parameters were given for the operation of the simulation model. Abbreviations of the parameters were defined by the authors. The definition of the parameters was made based on previous studies [37].

• Vv: Average speed of vehicle
• Vd: Average speed of drone
• Dv: Operating distance of vehicle
• Dd: Operating distance of drone, empty
• Dd2: Operating distance of drone, full
• Ev: Energy consumption of vehicle
• Ed: Energy consumption of drone, empty
• Ed2: Energy consumption of drone, full
• Cv: CO₂ emissions of van
• Cd: CO₂ emissions of drone, empty
• Cd2: CO₂ emissions of drone, full
• AVDv: Average delivery time of vehicle
• AVDd: Average delivery time of drone
• AVEv: Average energy consumption of vehicle/delivery
• AVEd: Average energy consumption of drone/delivery
• AVCOv: Average CO₂ emissions of vehicle/delivery
• AVCOd: Average CO₂ emissions of drone/delivery
• Emaxv: Total energy consumption of vehicle
• Emaxd: Total energy consumption of drone
• Cmaxv: Total CO₂ emissions of vehicle
• Cmaxd: Total CO₂ emissions of drone

Table 1. Factory parameters of simulated vehicle and drone, source: authors’ own work.

Factory Parameter	Fiat Ducatio BEV 3.5t LH2 122LE	DJI Matrice 300 RTK
Average speed (WTLP)	46.60 km/h	16.60 km/h
Operating distance (empty)	357.00 km	15.00 km
Operating distance (full)	317.00 km	8.45 km
Fuel tank/battery	79.00 kWh	548.00 Wh
Consumption (empty)	221.29 Wh/km	36.54 Wh/km
Consumption (full)	249.21 Wh/km	64.00 Wh/km
Weight delivered (full)	764.1 kg	2.70 kg
CO2 emissions (empty)	0.083 kgCO2/km	0.014 kgCO2/km
CO2 emissions (full)	0.094 kgCO2/km	0.024 kgCO2/km

Table 1. Factory parameters of simulated vehicle and drone, source: authors’ own work.

Factory Parameter	Fiat Ducatio BEV 3.5t LH2 122LE	DJI Matrice 300 RTK
Average speed (WTLP)	46.60 km/h	16.60 km/h
Operating distance (empty)	357.00 km	15.00 km
Operating distance (full)	317.00 km	8.45 km
Fuel tank/battery	79.00 kWh	548.00 Wh
Consumption (empty)	221.29 Wh/km	36.54 Wh/km
Consumption (full)	249.21 Wh/km	64.00 Wh/km
Weight delivered (full)	764.1 kg	2.70 kg
CO2 emissions (empty)	0.083 kgCO2/km	0.014 kgCO2/km
CO2 emissions (full)	0.094 kgCO2/km	0.024 kgCO2/km

2.4. Limitations

Goodchild and Toy (2018) investigated vehicle and drone cooperation in an instant delivery scenario [44]. They identified as a limitation of their study the estimation of the UAV energy consumption, as the energy consumption assumed in the scenario is considered too stringent, and weather aspects such as wind and temperature can also influence the energy consumption. In addition, adverse weather conditions may affect a number of other efficiency factors [50,51]. In adverse weather conditions, the drone may not be able to deliver packages; therefore, the presence of ground backup vehicles may be necessary [52]. Hwang et al. (2018) did not include detailed data on frequency and density of parcel delivery, but whether different social and economic conditions could improve the model’s performance [52]. An additional limitation of their study is that they did not consider the energy consumed by on-board electronic devices in the context of energy use. Drones have limited operating hours; their flight time is also affected by the type of load and the weight of the cargo carried, as heavier cargo reduces the flight time [53,54]. Safety factors should also be considered in drone research, such as the limitations of drones sharing the same airspace [55].

3. Results

The next section contains the results of the simulation models. Scenario (A) will be described first, followed by Scenario (B).

The following simulation models are the first steps of a more detailed research, and as such, they contain several constraints which serve as limitations at this stage. The average speed within city conditions, operation distance, energy consumption and CO₂ emission data of both tools are used in the simulation based on the factory parameters of the vehicle and the drone. In the case of vehicle traffic constraints such as traffic jams or a higher or lower traffic frequency (based on the time of day) can also alter average speed and delivery time. For this, an interval in speed was introduced.

The operating distance of the drone also restricts how far deliveries can be transported. Weather conditions and different environmental and surface factors can also alter the speed and delivery time of the drone as well, such as the weight of the cargo. Receiving the package by the customer in both cases and also refilling and recharging the drone or refueling the vehicle also take time, as does package loading. However, these factors are not yet incorporated at this stage. The simulation assumes that these conditions are currently optimal. During the calculation method, we calculated with the maximum capacity for all tested vehicles, and the transport process is standardized in all cases, so there is no traffic or other disturbing factors. The CO₂ emission and energy consumption data were determined based on the same calculation method, so data related to different transportation methods can be compared.

The following formulas were applied to calculate the average delivery time, CO₂ emissions and energy consumption of vehicles in both scenarios.

Average delivery time of vehicle (v) and drone (d):

$$\mathrm{A}\mathrm{V}\mathrm{D}\mathrm{v}=\frac{{\sum }_{\mathrm{i}=1}^{\mathrm{n}}\frac{{\mathrm{D}\mathrm{v}}_{\mathrm{i}}}{{\mathrm{V}\mathrm{v}}_{\mathrm{i}}}}{\mathrm{n}} \mathrm{A}\mathrm{V}\mathrm{D}\mathrm{d}=\frac{{\sum }_{\mathrm{i}=1}^{\mathrm{n}}\frac{{\mathrm{D}\mathrm{d}}_{\mathrm{i}}}{{\mathrm{V}\mathrm{d}}_{\mathrm{i}}}}{\mathrm{n}}$$

(1)

The average delivery time is based on the individual factory-provided parameters of the chosen vehicle and drone type, using their average speed and operating distance parameters.

Average CO₂ emission of vehicle (v) and drone (d):

$$\mathrm{A}\mathrm{V}\mathrm{C}\mathrm{O}\mathrm{v}=\frac{\sum _{\mathrm{i}=1}^{\mathrm{n}}{\mathrm{D}\mathrm{v}}_{\mathrm{i}}\times {\mathrm{C}\mathrm{v}}_{\mathrm{i}}}{\mathrm{n}} \mathrm{A}\mathrm{V}\mathrm{C}\mathrm{O}\mathrm{d}=\frac{\sum _{\mathrm{i}=1}^{\mathrm{n}}{(\mathrm{D}\mathrm{d}}_{\mathrm{i}}\times {\mathrm{C}\mathrm{d}}_{\mathrm{i}}+{\mathrm{D}\mathrm{d}2}_{\mathrm{i}}\times {\mathrm{C}\mathrm{d}2}_{\mathrm{i}})}{\mathrm{n}}$$

(2)

The average CO₂ emission of the vehicle and drone is also based on the factory data and traveled distance, taking also into account the difference in the CO₂ emissions of the drones if they are loaded with a package or empty when travelling back to the station.

Average energy consumption of vehicle (v) and drone (d):

$$\mathrm{A}\mathrm{V}\mathrm{E}\mathrm{v}=\frac{\sum _{\mathrm{i}=1}^{\mathrm{n}}{\mathrm{D}\mathrm{v}}_{\mathrm{i}}\times {\mathrm{E}\mathrm{v}}_{\mathrm{i}}}{\mathrm{n}} \mathrm{A}\mathrm{V}\mathrm{E}\mathrm{d}=\frac{\sum _{\mathrm{i}=1}^{\mathrm{n}}{(\mathrm{D}\mathrm{d}}_{\mathrm{i}}\times {\mathrm{E}\mathrm{d}}_{\mathrm{i}}+{\mathrm{D}\mathrm{d}2}_{\mathrm{i}}\times {\mathrm{E}\mathrm{d}2}_{\mathrm{i}})}{\mathrm{n}}$$

(3)

The average energy consumption of the vehicle and drone is based on factory data as well as traveled distance, taking again into account the difference in the energy consumption of the drones if they are loaded with a package or empty when travelling back to the station.

3.1. Scenario (A) Simulation with Vehicle

In the first scenario, the last-mile delivery occurs only with the assistance of a ground vehicle, which can deliver the packages from customer to customer. The city was divided into five areas, where a total number of 150 deliveries occur in a single day. From the distribution center, the vehicle can visit each area one at a time, where each node represents ten customers within the given area.

Based on the traveling distance and the average speed of the vehicle, the simulation is able to calculate the delivery time, energy consumption and CO₂ emissions by using the factory parameters of the vehicle (Fiat Ducatio BEV 3.5t LH2 122LE, which is a common vehicle used in last-mile delivery services). After the simulation, the average and total energy consumption and emission results are calculated.

Figure 3 shows the basic simulation model of Scenario (A) with the defined parameters.

After the simulation was completed, the data were gathered and summarized.

Table 2. Delivery statistics of a vehicle in Scenario (A), source: authors’ own work.

	Average Distance	Average Delivery Time	Total CO2 Emission	Total Energy Consumption
Area 1	2.67 km	0.10 h	7.54 kgCO2	19.99 kWh
Area 2	2.64 km	0.11 h	7.43 kgCO2	19.70 kWh
Area 3	2.78 km	0.10 h	7.83 kgCO2	20.76 kWh
Area 4	2.24 km	0.09 h	6.31 kgCO2	16.73 kWh
Area 5	2.61 km	0.07 h	6.25 kgCO2	16.56 kWh

contains the collected information in each area of the city. On average, the distance was 2.53 km between the customers, and the average delivery time was 12 min per customer. The average CO₂ emission per area was 7.07 kgCO₂ and the average energy consumption per area was 18.75 kWh.

As we can see in

Table 3. Summarized delivery, energy consumption and emission statistics in Scenario (A), source: authors’ own work.

	Category	Result
1.	Total distance	379.77 km
2.	Average distance	2.53 km
3.	Average delivery time	0.12 h
4.	Total CO2 emission	35.70 kgCO2
5.	Total energy consumption	94.64 kWh

, after the simulation, a total distance of 379.77 km was traveled to deliver all packages to the customers. The total energy consumption of the vehicle was 94.64 kWh and the total CO₂ emission was 35.7 kg CO₂ based on the results and the factory parameters.

3.2. Scenario (B) Simulation with Vehicle and Drone

In the second scenario, the last-mile delivery is executed by both the assistance of a vehicle and drone, resulting in a combined delivery where the vehicle only transports the packages to the drone stations located in the five areas of the city. In this scenario, a total number of 150 deliveries occur in a single day. From the distribution center, the vehicle can visit each area one at a time, delivering the packages, and then the drones transport the packages to the customers. Each node represents ten customers within the given area. After the drone stations receive the packages by the delivery cars, they deliver the packages to the customers and then return to their station via the same path.

In this scenario, choosing the right location of the drone stations is crucial for an effective and successful delivery service. At this point, the simulation assumes that ideal locations were chosen for the stations to cover the city, but in real life, such locations could be challenging to obtain and it is certain that trade-offs should be made. The simulation also considers the location choice as a limitation of the current simulation.

Based on the traveling distance and the average speed of the vehicle and the drones, the simulation is able to calculate the delivery time, energy consumption and CO₂ emissions by using the factory parameters, and after the simulation, the average and total energy consumption and emission results are calculated again.

Figure 4 shows the basic simulation model of Scenario (B) with the defined parameters.

After the simulation was completed, the data were gathered and summarized.

Table 4. Delivery statistics of a vehicle in Scenario (B), source: authors’ own work.

	Distance	Delivery Time	CO2 Emission	Energy Consumption
Area 1	13.58 km	0.49 h	1.28 kgCO2	3.39 kWh
Area 2	10.74 km	0.31 h	1.01 kgCO2	2.68 kWh
Area 3	10.07 km	0.61 h	0.95 kgCO2	2.51 kWh
Area 4	13.71 km	0.43 h	1.29 kgCO2	3.42 kWh
Area 5	9.30 km	0.26 h	0.87 kgCO2	2.32 kWh

contains the collected information regarding the vehicle in each area of the city. On average, the distance was 11.48 km between the drone stations and the average delivery time was 26 min per drone station. The average CO₂ emission per area was 1.08 kgCO₂ and the average energy consumption per area was 2.86 kW.

After the packages are delivered to the drone stations, the drones deliver the packages to the customers within that area.

Table 5. Delivery statistics of drones in Scenario (B), source: authors’ own work.

	Average Distance	Average Delivery Time	Total CO2 Emission	Total Energy Consumption
Drone Station 1	4.09 km	0.49 h	2.33 kgCO2	6.17 kWh
Drone Station 2	4.47 km	0.54 h	2.55 kgCO2	6.74 kWh
Drone Station 3	4.03 km	0.49 h	2.30 kgCO2	6.08 kWh
Drone Station 4	5.26 km	0.63 h	3.00 kgCO2	7.94 kWh
Drone Station 5	7.19 km	0.87 h	2.83 kgCO2	7.48 kWh

shows that the average travel distance of a delivery was 5.01 km, while the average delivery time per customer was 36 min. The average CO₂ emission per station was 2.6 kgCO₂ and the average energy consumption per area was 6.88 kWh.

Table 6. Summarized delivery, consumption and emission statistics in Scenario (B), source: authors’ own work.

	Vehicle	Drone	SUM
Total distance	57.40 km	732.93 km
Average distance	11.48 km	4.89 km
Average delivery time	0.42 h	0.71 h
Total CO2 emission	5.40 kgCO2	13.93 kgCO2	19.32 kgCO2
Total energy consumption	14.30 kWh	36.84 kWh	51.15 kWh

contains the combined information of the second scenario regarding the support vehicle and the drones which provide the last-mile delivery service.

In this scenario, it can be clearly seen that the total CO₂ emission of the vehicle is only 5.4 kgCO₂ compared to the 94.64 kgCO₂ from Scenario (B). In this scenario, the vehicle and drones’ combined CO₂ emission was 19.32 kgCO₂ and the combined energy consumption was 51.15 kWh.

According to the results of the simulation presented in

Table 7. Comparison of Scenario (A) and Scenario (B) emission and energy consumption results.

	Scenario 1	Scenario 2	Difference
Total CO2 emission	35.70 kgCO2	19.32 kgCO2	16.38 kgCO2
Total energy consumption	94.64 kWh	51.15 kWh	43.49 kWh

, a drone-assisted combined last-mile delivery system seen in Scenario (B) can reduce the total CO₂ emission of the service by 16.38 kgCO₂, while the total energy consumption can also be reduced by 43.49 kWh under the circumstances of the simulation. While these results are obtained with certain limitations regarding the environmental and technological factors of the delivery service, they are promising from the sustainability perspective. By incorporating these details into a more refined future version of the simulation, we would be able to determine whether such a development is worthwhile or not on a larger scale from a last-mile delivery company’s point of view.

4. Discussion

Last-mile delivery services face a number of challenges that can make it difficult to deliver packages efficiently and cost-effectively. One of the main challenges is the increasing demand for faster and more flexible delivery options, which can be difficult to realize in congested urban areas. The high cost of transportation and logistics can also be a challenge for last-mile delivery services, particularly for small and medium-sized businesses. Weather conditions, traffic congestion and other external factors can also impact the efficiency and reliability of last-mile delivery services. In addition, last-mile delivery services must be able to adapt to changing consumer expectations and preferences, which can be challenging in a rapidly evolving industry.

One of the main challenges facing last-mile delivery services is the high level of emissions generated by traditional delivery methods. The use of fossil-fuel-powered vehicles for last-mile delivery can contribute to air pollution and greenhouse gas emissions, which can have significant environmental impacts. Emissions from last-mile delivery vehicles can also contribute to local air quality issues, particularly in urban areas with high levels of congestion. The high fuel consumption of traditional delivery vehicles can also be a challenge, as it can lead to increased operating costs for delivery companies. In addition, the volatility of fuel prices can make it difficult for delivery companies to predict and manage their expenses effectively. The limited range of some electric delivery vehicles can also be a challenge for last-mile delivery services, as they may not be able to travel as far or carry as much cargo as traditional vehicles.

Drones have the potential to significantly reduce the energy consumption and emissions associated with last-mile delivery services. Traditional delivery vehicles such as vans and trucks consume large quantities of fuel and produce a large amount of carbon emissions, contributing to air pollution and climate change. In contrast, drones require significantly less energy to operate, as they are powered by lightweight batteries and do not need to carry heavy loads. The use of drones for last-mile delivery can also help reduce traffic congestion and emissions in urban areas, as they are able to fly over obstacles and reach their destination more quickly. Overall, the adoption of drone technology in last-mile delivery services can lead to significant reductions in energy consumption and emissions, making it an attractive option for companies seeking to reduce their carbon footprint and contribute to a more sustainable future.

According to the results of the two simulations, within the defined parameters and limitations, a combined system of drones and vehicles can result in a last-mile delivery system which produces less emissions and consumes less energy while providing the same level of service.

However, despite the many advantages of using drones for last-mile delivery, there are still several obstacles and disadvantages that need to be addressed. One of the main obstacles is regulatory barriers, as many countries and cities have strict regulations with regard to the use of drones for commercial purposes. Another challenge is the limited range and payload capacity of drones, which can make it difficult to deliver larger or heavier packages. Drones are also vulnerable to bad weather conditions such as high winds, rain and snow, which can affect their ability to fly and deliver packages.

The high cost of drone technology and the need for specialized equipment can also be a barrier for smaller businesses or companies with limited resources. Security and privacy concerns are also a major disadvantage of using drones for delivery, as they can be susceptible to hacking or theft. The noise produced by drones during operation can also be a nuisance to people living in residential areas, which can lead to complaints and opposition to the technology. The need for dedicated landing zones and the potential for drones to collide with other objects or people can also be a safety concern.

The limited range of drones can also be a disadvantage, as they typically have a range of only a few kilometers, which can limit their use for longer distances. To extend the range of drones, additional charging stations or larger batteries may be required, which can add to the overall cost. The range and payload capacity of drones can also be affected by weather conditions, which can further limit their use in certain areas or at certain times. The cost of infrastructure such as landing zones, charging stations and maintenance facilities can also add to the overall cost of using drones for last-mile delivery.

Companies may also need to invest in specialized training for their employees to operate and maintain the drones, which can add to the cost. Another cost disadvantage of drones is that they may require ongoing updates and upgrades to their technology, which can be expensive. Overall, while the use of drones for last-mile delivery services offers many benefits, the cost and range limitations associated with this technology can be a significant disadvantage for some businesses.

The presented results were calculated considering ideal conditions, but in reality, not everything is ideal. As an extension of our research, we would like to point out that this is a very sensitive system, with a high number of bound parameters, which, if changed, may lead to different results in the current simulation, which can lead to a less favorable drone-assisted last-mile delivery scenario. This phenomenon can be illustrated with the two graphs in Figure 5.

If the length of the last-mile delivery is changed due to the suboptimal location of drone stations or due to avoidance of landmarks, this can significantly affect range, efficiency and CO₂ emissions. Aside from these parameters, other aspects may change, such as weather conditions, the type of drones used, delivery circumstances, or other delivery companies that may use drones and whose drone movements may interfere with one another.

In addition to all this, it is also important to take into account costs, which are significantly influenced and affected by changes in parameters. For example, the cost of transshipment, system maintenance and upkeep when using drones can make a significant difference between delivery with or without drone assistance.

5. Conclusions

Drones are increasingly being used in the last-mile delivery of goods, providing a faster and more efficient alternative to traditional delivery methods. With the ability to bypass traffic and roadblocks, drones can reach their destination in a shorter amount of time, making them ideal for delivering urgent packages. Delivery drones are also able to operate in remote areas or locations with limited access, making them a great option for delivering essential supplies to people in need.

The aim of the research was to examine the advantages and disadvantages of the application of drones in last-mile delivery systems from a sustainability point of view regarding CO₂ emissions and energy consumption. Discrete event simulation was applied to examine how the assistance of drones in parcel delivery services can influence the sustainability of such services from a CO₂ emission and energy consumption perspective in an urban environment, while also pointing out the possible challenges of such drone-assisted systems for delivery companies. Two scenarios were created and compared with each other. In the first scenario, a regular vehicle-based delivery system is considered, while in the second scenario, the last-mile delivery system is also assisted by drones.

According to the results of the simulations, within the defined parameters, a combined system of drones and vehicles can result in a last-mile delivery system which has less CO₂ emissions and consumes less energy, while providing the same level of service. As the results of the simulations show, in the first scenario without drones, the total CO₂ emission was 35.7 kgCO₂ and the total energy consumption was 94.64 kWh, while in the second scenario using a drone-assisted system, the total CO₂ emission was 19.32 kgCO₂ and the total energy consumption was 51.15 kWh, resulting in a difference of 16.38 kgCO₂ in emission and a 43.49 kWh in energy consumption. It can be seen that the use of drones in last-mile delivery can reduce the carbon footprint of logistics companies, as they require less fuel and produce less CO₂ emissions than traditional delivery vehicles.

However, despite the possible advantages of using drones for last-mile delivery, there are also several obstacles and disadvantages such as regulatory barriers, the limited range and payload capacity of drones, or weather and surface conditions. The high cost of drone technology and the need for specialized equipment and training can also be a barrier for smaller businesses or companies with limited resources.

While these results were obtained with certain limitations regarding the environmental and technological factors of the delivery service, they are promising from a sustainability perspective. If the length of the last-mile delivery is changed due to the suboptimal location of drone stations or due to unavoidable delivery conditions, the range, efficiency and CO₂ emissions can be significantly affected. The cost of the overall service under different conditions is a very important aspect, as it can be significantly influenced and affected by changes in parameters. By incorporating these details into a more refined future version of the simulation, we would be able to determine whether such a development is worthwhile on a larger scale from a last-mile delivery company’s point of view.

As the technology continues to improve and regulations become more accommodating, the use of drones in last-mile delivery services is expected to become increasingly common in the coming years.

This study presents the initial steps of the research. Further research is needed to incorporate more environmental, technological and also financial factors to determine whether such a development is worthwhile on a larger scale from a last-mile delivery company’s point of view.