Available via license: CC BY 4.0
Content may be subject to copyright.
Citation: Riaventin, V.N.; Cakravastia,
A.; Cahyono, R.T.; Suprayogi.
Sustainable Synchronization of Truck
Arrival and Yard Crane Scheduling in
Container Terminals: An Agent-Based
Simulation of Centralized and
Decentralized Approaches
Considering Carbon Emissions.
Sustainability 2024,16, 9743. https://
doi.org/10.3390/su16229743
Academic Editors: Jozef Gašparík and
Davor Dujak
Received: 25 September 2024
Revised: 21 October 2024
Accepted: 1 November 2024
Published: 8 November 2024
Copyright: © 2024 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Article
Sustainable Synchronization of Truck Arrival and Yard Crane
Scheduling in Container Terminals: An Agent-Based Simulation
of Centralized and Decentralized Approaches Considering
Carbon Emissions
Veterina Nosadila Riaventin 1,2,3,* , Andi Cakravastia 1,2,3, Rully Tri Cahyono 1,2,3 and Suprayogi 1,2,3
1Industrial Engineering Program, Faculty of Industrial Technology, Bandung Institute of Technology,
Bandung 40132, Indonesia; andi@itb.ac.id (A.C.); r.t.cahyono@itb.ac.id (R.T.C.); yogi@itb.ac.id (S.)
2Research Group of Industrial System and Techno-Economics, Bandung Institute of Technology,
Bandung 40132, Indonesia
3Center for Logistics and Supply Chain Studies, Bandung Institute of Technology, Bandung 40132, Indonesia
*Correspondence: veterina@itb.ac.id
Abstract: Background: Container terminal congestion is often measured by the average turnaround
time for external trucks. Reducing the average turnaround time can be resolved by controlling the
yard crane operation and the arrival times of external trucks (truck appointment system). Because
the truck appointment system and yard crane scheduling problem are closely interconnected, this
research investigates synchronization between the approaches used in truck appointment systems
and yard crane scheduling strategies. Rubber-tired gantry (RTG) operators for yard crane scheduling
operations strive to reduce RTG movement time as part of the container retrieval service. However,
there is a conflict between individual agent goals. While seeking to minimize truck turnaround
time, RTGs may travel long distances, ultimately slowing down the RTG service. Methods: We
address a method that balances individual agent goals while also considering the collective objective,
thereby minimizing turnaround time. An agent-based simulation is proposed to simulate scenarios
for yard crane scheduling strategies and truck appointment system approaches, which are centralized
and decentralized. This study explores the combined effects of different yard scheduling strategies
and truck appointment procedures on performance indicators. Various configurations of the truck
appointment system and yard scheduling strategies are modeled to investigate how those factors
affect the average turnaround time, yard crane utilization, and CO
2
emissions. Results: At all levels
of truck arrival rates, the nearest-truck-first-served (NTFS) scenario tends to provide lower external
truck turnaround times than the first-come-first-served (FCFS) and nearest-truck longest-waiting-time
first-served (NLFS) scenario. Conclusions: The decentralized truck appointment system (DTAS)
generally shows slightly higher efficiency in emission reduction compared with centralized truck
appointment system (CTAS), especially at moderate to high truck arrival rates. The decentralized
approach of the truck appointment system should be accompanied by the yard scheduling strategy
to obtain better performance indicators.
Keywords: agent-based simulation; decentralized; carbon emissions; synchronization; truck appoint-
ment system; yard scheduling strategy
1. Introduction
Ports are vital to the global logistics network, handling approximately 80% of interna-
tional trade by volume [
1
] and playing a distinct role in promoting sustainability within
the shipping industry [
2
]. Recognizing sustainability as a global concern, the United Na-
tions (UN) introduced the sustainable development goals (SDGs) in 2015, setting 17 goals,
169 targets
, and 230 indicators to drive global change by 2030. The International Maritime
Organization (IMO) has highlighted the significance of the shipping industry in achieving
Sustainability 2024,16, 9743. https://doi.org/10.3390/su16229743 https://www.mdpi.com/journal/sustainability
Sustainability 2024,16, 9743 2 of 25
all SDGs by facilitating global trade and economic growth [
3
]. As hubs of activity and key
economic centers, ports contribute significantly to the global SDG agenda. Sciberras and
Silva [
3
] provided a UNSDG-based sustainability framework for ports, addressing actions
aligned with specific SDGs. Several sustainability actions related to drayage operation are
reducing related traffic congestion (SDG 11: sustainable cities and communities), optimiz-
ing freight traffic (SDG 9: industry innovation and infrastructure and SDG 12: responsible
consumption and production), and controlling pollution through all activities (SDG 6: clean
water and sanitation).
However, despite these sustainability efforts in the drayage operation, real-world
conditions present significant challenges. Many ports experience significant delays in
container transportation due to capacity constraints in both ports and road networks,
leading to congestion. The congestion that occurs both inside and outside the port causes
various losses, including a decline in productivity in container terminal operations, high
logistics costs, longer truck operating times, and increased pollution (fuel emissions). Chen
et al. [
4
] stated that congestion causes emissions from idling trucks. Idle trucks operate
inefficiently, with an energy efficiency level of around 3% compared with trucks that operate
at 40% energy efficiency. Idle trucks produce high levels of NOx and particulate matter
(PM), both of which are substances that significantly contribute to air pollution.
A key factor contributing to terminal congestion is the fluctuating arrival of exter-
nal trucks, which can result in mismatches between resource constraints and container
transportation activities [
5
]. Shipping companies own external trucks that move containers
between depots and terminals [
6
]. When container transportation activity exceeds resources,
turnaround times are delayed, and when resources exceed container transportation activity,
resources are wasted [5].
Port congestion causes significant losses for container terminals, shipping companies,
and consignees. For container terminals, severe congestion can delay or even halt yard
operations as yard cranes struggle to move among the many external trucks, making these
terminals less attractive to shipping companies seeking efficient pick-up and delivery
services [
7
]. Shipping companies also incur losses due to port congestion, facing increased
shipping times and higher costs for loading and unloading activities, and may need to
reroute ships, further degrading service quality [
8
]. Consignees and trucking companies
experience delays in pick-up and delivery times due to high congestion levels, leading
to increased fuel and labor costs, and container shortages can drive up rental costs when
demand outpaces supply [
9
]. Port congestion is often measured by the average turnaround
time for external trucks, which encompasses the period from a truck’s entry to its exit after
completing a container transaction. Increased turnaround times indicate higher congestion,
impacting terminal efficiency and trucking company performance [6]. Congestion at both
the gate and in the yard causes high truck turnaround times within the terminal [
10
].
Reducing congestion can significantly shorten these turnaround times, leading to improved
customer satisfaction [
11
]. Researchers in [
12
] describe the cost of yard congestion in terms
of truck turnaround times. When yard cranes are occupied due to congestion in yard
blocks, this results in longer turnaround times for external trucks being serviced in those
areas. There is a clear connection between available time for the container terminal and
the trucking company with container terminal space related to truck densities and storage
spaces, as coordinating the arrival of external trucks can help reduce congestion in the
terminal’s storage areas [6].
To manage growing container volumes, ports must achieve faster turnaround times
and greater container throughput, which refers to the efficiency and speed of moving
goods through the transport chain [
11
]. The efficiency of a container terminal is typically
evaluated by two key metrics: (1) the duration of vessel berthing and (2) the turnaround
time for external trucks [
6
]. Reducing truck turnaround time in terminals is a critical
concern for container terminals, trucking companies, and government regulators [10].
Truck turnaround time is derived from the sum of the truck waiting time and the truck
service time. Waiting time is the idle time of the truck before being serviced; service time
Sustainability 2024,16, 9743 3 of 25
is the time required to move containers. Increased waiting time reduces the operational
efficiency of the terminal’s stacking yard [
13
]. The depiction of the activity flow considered
in truck turnaround time can be seen in Figure 1.
Figure 1. External truck turnaround time in container terminal.
Due to the limited land area available for expansion around most ports, terminal opera-
tors must find ways to enhance productivity using existing resources [
11
].
Two potential solutions are as follows: (1) expediting container service times and
(2) regulating the arrival times of external trucks. Numerous studies have examined
these methods to reduce average truck turnaround times. The container service time can be
reduced by deploying more yard cranes and minimizing the frequency of container reloca-
tions [
14
,
15
]. Huynn et al. [
14
] employed a simulation to assess how many additional yard
cranes would be needed to meet the desired turnaround time. Their findings demonstrated
that increasing the number of cranes can substantially decrease the average turnaround
time. Additionally, container relocations affect service times; when trucks arrive, contain-
ers may need to be shifted if stacked underneath others. Boysen and Emde [
15
] explore
solutions to these challenges using operational research techniques to speed up the service
time for external trucks. They suggest that implementing a truck appointment system is
an efficient way to manage and control truck arrival times. This system allows for more
efficient scheduling of container pick-up activities.
It is important to recognize that the truck arrival process and container handling
operations are interconnected. Minimizing the truck waiting time in the yard, which
consumes the largest share of turnaround time, is essential for reducing overall turnaround
time [
16
]. Reducing turnaround time can be beneficial for alleviating congestion, optimizing
freight traffic, and reducing pollution, which ultimately contributes to achieving the SDG
targets for 2030, particularly SDG 6, 9, 11, and 12. Therefore, it is valuable to study a
coordinating approach to optimize yard crane working time with truck appointments [
17
].
Trucks queue at terminal gates until yard cranes are available to serve them, and, once in
the yard, turnaround time depends on both the truck-to-crane ratio and the yard crane
service strategy [
18
]. Effective synchronization between the gate and yard is essential for
designing an efficient truck appointment system [19].
External truck congestion at ports primarily stems from issues related to resource
allocation and gate scheduling. Although tracking real-time queue lengths at gates can
help in developing appointment systems, the actual port operations involve a complex
interplay of three main subsystems: (1) berth, (2) yard, and (3) gate system [
19
]. In
this study, we concentrate on these interrelationships and thoroughly investigate the
synchronization of truck arrival scheduling with the yard crane scheduling strategy in the
container terminal yard. The authors examined yard scheduling strategies in combination
with both centralized and decentralized truck appointment procedures, assessing how
these combinations impact terminal performance. To clarify the specific contributions of
this study, the following key areas were addressed:
1.
Comparison of yard scheduling strategies: compared existing yard crane scheduling
strategies, specifically focusing on FCFS (first-come-first-served) and the proposed
Sustainability 2024,16, 9743 4 of 25
NTFS (nearest-truck-first-served) and nearest-truck longest-waiting-time first-served
(NLFS), to assess operational efficiencies.
2.
Comparison of centralized and decentralized truck appointment procedures: eval-
uated the impact of centralized vs. decentralized procedures on truck turnaround
times and yard efficiency.
3.
Performance assessment of combined strategies: assessed the effect of combining
different yard scheduling strategies with both appointment systems to understand
their influence on terminal performance.
4.
Carbon emissions analysis: measured carbon emissions across different combinations
of yard scheduling and truck appointment strategies to identify environmentally
efficient practices.
2. Related Works
This study investigated the interrelationship between the procedures used in the truck
appointment system and the yard scheduling strategy. Therefore, this literature review
examines extant research into yard scheduling strategy and truck appointment systems,
and research that combines these two methods, namely the integrated truck appointment
system with yard crane scheduling strategy.
2.1. Yard Scheduling Strategy
Enhancing container service times can be accomplished by augmenting the quantity
of cranes deployed within the storage yard and minimizing the need for container reloca-
tions [
14
,
15
]. In [
14
], simulation methods were employed to determine how many extra
yard cranes would be required to reach the target average turnaround time for external
trucks. Their research indicated that increasing the number of cranes significantly reduces
the average turnaround time for external trucks. Adding more yard cranes to reduce truck
turn times may appear to be a straightforward solution for terminals that stack containers.
However, the significant initial investment required to install additional cranes, along
with the increased ongoing maintenance and operating expenses, poses significant barriers
to this seemingly obvious solution. Seeking alternative solutions, Kim et al. [
16
] tested
various sequencing methods to deploy yard cranes more effectively, thereby reducing the
in-terminal wait time of trucks. Huynh and Vidal [
18
] employed an agent-based simulation
to examine a specific aspect of the issue: the service strategies of yard cranes. Their goal
was to identify the optimal strategies for minimizing truck wait times, considering the
random arrival of trucks.
One issue that prolongs service time for external trucks is container relocation. When
picking up or dropping off containers, external trucks often encounter containers that are
stacked beneath others, necessitating the relocation of the upper containers first. Boysen
and Emde [
15
] addressed this problem using operational research methods to expedite
service times for external trucks. Wasesa et al. [
20
] proposed a synchronization policy that
assigns the requests of the incoming trucks to the topmost container. They found that
achieving a 100% utilization rate of the synchronization policy will reduce the probability
of relocation to less than 1%.
Diverging from the approach in [
18
], this study instead synchronizes yard scheduling
and truck appointment systems, focusing on the yard scheduling model that aims to
minimize truck turnaround time, that is, the shared performance indicator for each problem
involved, instead of other performance indicators for yard scheduling problems.
2.2. Truck Appointment System
In the operational decision-making process, a truck appointment system considers the
available time for container terminal resources and trucking company resources (such as
trucks and drivers) as well as the terminal’s capacity in terms of truck density and storage
space [
6
]. The review study written by the author in [
19
] describes the truck appointment
system as a two-factor decision-making framework focused on time and space. Its objective
Sustainability 2024,16, 9743 5 of 25
is to balance the flow of trucks arriving at the gate throughout the day, thereby reducing
the strain on the terminal’s physical capacity.
Relevant truck appointment systems can be categorized based on various methods,
such as (1) determining reservation quotas, (2) scheduling and optimizing truck arrivals,
(3) scheduling and optimizing truck arrivals with collaboration, and (4) minimizing empty
truck trips. Research studies can also be organized based on the approaches they utilize,
such as queuing theory, mathematical modeling, simulations, or a mix of these techniques,
as demonstrated in a variety of the academic literature.
The first method, vessel-dependent time windows (VDTWs), sets a limit on the vol-
ume of trucks permitted to arrive within each designated time window. Chen et al. [
21
]
introduced the VDTW method to control truck arrivals by categorizing trucks and allo-
cating them to different time windows. Gracia et al. [
22
] analyzed the effects of gradually
introducing a truck appointment system and a dedicated entry lane approach for external
trucks on waiting times through simulations. Their findings indicated that increasing the
number of trucks using the appointment system led to a reduction in average waiting
times. Huynh [
23
] investigated the effect of truck quotas on average turnaround times,
and showed that optimal truck quotas can significantly lower turnaround times. Torkjazi
et al. [
24
] proposed a method to evenly spread truck arrivals across the day to prevent
congestion at gates and yards, while also taking into account drayage truck routes.
The second approach focuses on coordinating truck arrivals by balancing the objectives
and constraints of the trucking companies and the container terminal. This system considers
the available resources and terminal capacity regarding truck density and storage space [
6
].
Huiyun et al. [
19
] described this method as a two-dimensional decision-making system
that balances truck arrivals over time to reduce spatial pressure at the terminal. Truck
scheduling entails selecting time slots for truck arrivals, which can be determined by the
trucking company and the port [25–27].
The third approach involves cooperation and negotiation between trucking companies
and terminal operators to determine vehicle arrival timings and manage operations during
high-demand periods [
25
]. Phan and Kim [
28
] introduced a decentralized decision-making
model to facilitate these negotiations. Azab et al. [
12
] proposed a dynamic collaboration
model integrating discrete-event simulation with mixed integer programming. Wasesa
et al. [
29
] proposed an overbooking reservation mechanism (ORM) and conducted agent-
based simulations to evaluate the ORM’s performance. Research indicates that using ORMs
can result in high productivity and service levels while minimizing negative externalities
such as long queues, overtime, and greenhouse gas emissions.
The fourth method focuses on minimizing empty trips by grouping the delivery of
export containers and the pick-up of import containers. Islam et al. [
30
] were among the pio-
neers in addressing the issue of empty container trucks, proposing a dynamic truck-sharing
system to allocate export containers to vacant slots in empty trucks. Schulte et al. [
31
]
developed a truck appointment model that considers the impact of empty trips on carbon
emissions and costs, using an optimization model based on multiple traveling salesman
problems with time windows. Caballini et al. [
32
] combined optimization techniques and
data analysis methods to minimize the occurrence of empty truck journeys and improve
service levels using a truck appointment system. Dekker et al. [
33
] investigated the idea of
a chassis exchange terminal, an off-dock facility where truck drivers can swap chassis to
alleviate congestion.
Research on truck appointment systems often uses mathematical modeling, queuing
theory, simulations, or a combination of these approaches. Phan and Kim [
25
,
28
] employed
mathematical modeling to tackle the issues in their studies; Huynh [
23
] and Murty et al. [
34
]
relied on simulation techniques; Zhang et al. [
10
] and Chen et al. [
4
] addressed the problem
through the queuing theory; and Azab et al. [
12
] introduced a method that combines a
mixed-integer programming (MIP) model with a simulation model to minimize turnaround
times for external trucks and mitigate disruptions caused by adjusting truck arrivals away
from the preferred times of the trucking companies.
Sustainability 2024,16, 9743 6 of 25
Considering the goals of this study, the scheduling problem for truck arrivals can
focus on maximizing resource utilization while minimizing transportation costs, truck
turnaround times, waiting times, empty trips, and emissions, among other factors. Huynh
and Walton [
5
] investigated how restricting truck arrivals affects both truck turnaround
times and crane usage. Chen and Yang [
27
] developed an optimization model for managing
export container truck operations, aiming to minimize the overall cost of export container
activities. Zhang et al. [
10
] proposed a stationary-based queuing optimization model for
truck appointments to alleviate severe congestion in container terminals; their method
demonstrated high accuracy in predicting queue lengths at both the gates and the yard, and
effectively reduced truck turnaround times. Azab et al. [
12
] use the average turnaround
time of external trucks to estimate congestion costs; this approach is considered more
realistic and comprehensive because it considers both waiting and service times at the gate.
As shown in Table 1, previous research has commonly utilized mathematical modeling,
operational research, or discrete-event simulation approaches. However, these methods
often neglect the communication and operational sequence aspects of the appointment
reservation process in their formulation and evaluation of improvement proposals or mod-
els. In contrast, this study employs business process analysis and agent-based simulation
techniques to analyze, propose, and evaluate the proposed solution. Unlike prior research
conducted by authors in [
29
], which used agent-based modeling for the truck appoint-
ment system, this research investigates various configurations of the truck appointment
system and yard scheduling strategies to assess how those factors affect average truck
turnaround time.
Table 1. Truck appointment system literature review.
Author(s)
Controlling Arrival Times Methodology
Quota Truck
Scheduling
Collaboration
Scheduling
Combine
Export
Import
Container
Mathematical
Modeling Simulation
Model
Queueing
System
Exact
Approximation
Huynh et al. (2004) [14]✓
Huynh (2009) [23]✓ ✓ ✓
Chen and Yang
(2010) [27]✓ ✓ ✓
Chen et al. (2013) [21]✓ ✓ ✓
Islam et al. (2013) [30]✓Business Process Re-engineering
Phan and Kim
(2015) [28]✓ ✓ ✓
Boysen and Emde
(2016) [15]✓
Phan and Kim
(2016) [25]✓ ✓ ✓
Gracia et al. (2017) [22]✓ ✓
Schulte et al. (2017) [31]✓ ✓
Torkjazi et al. (2018) [
24
]
✓ ✓
Riaventin and Kim
(2019) [26]✓ ✓
Yi et al. (2019) [35]✓ ✓
Azab et al. (2020) [12]✓ ✓ ✓
Caballini et al.
(2020) [32]✓ ✓
Wasesa et al. (2021) [29]✓ ✓
Sustainability 2024,16, 9743 7 of 25
2.3. Synchronization of Truck Appointment System with Yard Crane Scheduling Strategy
A few studies on the design of container terminal operations that address both the
yard and gate areas in parallel are presented in Table 2. Guo et al. [
36
] aims to improve
the efficiency of yard crane operations by using predicted vehicle arrival information
to dynamically dispatch the crane. Hwang et al. [
37
] and Wasesa et al. [
20
] utilize the
container arrival information to improve container terminal’s performance in terms of
truck waiting time. Zhou et al. [
38
] proposed an integrated optimization method for
simultaneously determining yard crane schedules and vehicle parking positions, utilizing
Chebyshev’s movement to allow for the concurrent movement of the yard crane’s gantry
and trolley. Azab and Morita [
17
] presented a decision support system designed to schedule
optimized truck appointments, truck service orders, and container relocations within the
yard. They developed a bi-objective optimization model that considers container relocation
(terminal perspective) and appointment shifting (trucking company perspective). Gao
and Ge [
39
] proposed a model that addresses both the truck assignment and yard crane
routing problems simultaneously to optimize each yard crane’s servicing of the trucks
and to establish the ideal sequence for yard cranes’ handling of containers for each truck.
Talaat et al. [
40
] proposed a mixed-integer programming model to optimize the scheduling
of external trucks and yard cranes. The primary goals are to minimize CO
2
emissions,
reduce truck turnaround time, close the gap between trucking companies’ preferred and
appointed arrival times, and lower the energy consumption of yard cranes. Finally, Huang
et al. [
41
] proposed a unified model that addresses both the scheduling of trucks and empty
containers in container drayage operations, along with the appointment problem.
Table 2. Synchronization of truck appointment system and yard crane scheduling literature review.
Author Operation Area Truck App. System Yard Scheduling Method
Yard
Gate
Drayage
Decentralized Centralized
Seq. Strategy Opt. Heuristic DES ABS
Guo et al. (2011) [36]✓ ✓ ✓ ✓ ✓ ✓
Wasesa et al. (2012) [
20
]
✓ ✓ ✓ ✓ ✓
Hwang et al. (2019) [37]✓ ✓ ✓ ✓ ✓ ✓
Zhou et al. (2020) [38]✓ ✓ ✓ ✓ ✓ ✓
Azab and Morita
(2022) [17]✓ ✓ ✓ ✓ ✓
Ma et al. (2022) [42]✓ ✓ ✓ ✓ ✓ ✓
Gao and Ge (2023) [39]✓ ✓ ✓ ✓ ✓ ✓
Talaat et al. (2023) [40]✓ ✓ ✓ ✓ ✓
Huang et al. (2024) [
41
]
✓ ✓ ✓ ✓
This research ✓ ✓ ✓ ✓ ✓ ✓
Seq.—sequence, Opt.—optimization, DES—discrete-event simulation, ABS—agent-based simulation.
Based on the previous research in Table 2, this research makes several important con-
tributions that position it distinctly within the body of prior work on yard crane scheduling
and truck appointment systems. This research significantly contributes to the field by
exploring the simultaneous synchronization of yard crane scheduling and external truck
appointment systems. Unlike prior studies, such as Talaat et al. [
40
] and Ma et al. [
42
],
which approached yard crane and truck scheduling using multi-stage programming meth-
ods, this study investigates how the combination of yard crane scheduling and truck
appointments jointly influences truck turnaround time, offering a more comprehensive
solution to optimize terminal operations. Additionally, this research introduces and com-
pares both centralized and decentralized truck appointment systems, distinguishing itself
from previous studies like Wasesa et al. [
20
], Hwang et al. [
37
], Gao and Ge [
39
], and
Huang et al. [
41
], which primarily focused on centralized systems. By evaluating de-
centralized systems—where trucking companies have more control over appointment
Sustainability 2024,16, 9743 8 of 25
schedules—alongside centralized systems, this study highlights the increased flexibility of
decentralized approaches while maintaining or enhancing key performance metrics such
as truck turnaround time and yard crane utilization.
Furthermore, this research emphasizes the integration of yard crane scheduling strate-
gies with both centralized and decentralized truck appointment systems. While many
studies have addressed truck scheduling and yard crane sequencing, such as those by Guo
et al. [
36
], Zhou et al. [
38
], and Azab and Morita [
17
], this study focuses on sequencing
rules or strategies for yard crane scheduling, with a focus on feasibility in real-world sys-
tems. Unlike the works of Wasesa et al. [
20
] and Gao and Ge [
39
], which also considered
scheduling strategies, this research evaluates the combination of these strategies with both
decentralized and centralized truck appointment systems, offering a more comprehensive
analysis of how these elements interact to improve terminal efficiency.
To achieve this, this study employs agent-based simulation (ABS), in contrast with
the optimization models, heuristics, and discrete-event simulation (DES) methods used
by previous researchers such as Guo et al. [
36
], Zhou et al. [
38
], Gao and Ge [
39
], and
Hwang et al. [
37
]. ABS is particularly well suited for modeling dynamic interactions be-
tween agents, such as trucking companies and terminal operators, in yard crane operations
where each RTG operator works independently within specific blocks. This often creates
conflicts, as minimizing truck waiting times may lead to longer crane travel distances,
slowing down overall service. ABS captures these complexities, allowing for flexible
modeling of individual agent objectives—balancing truck waiting times with crane move-
ment efficiency—while addressing the broader goal of minimizing truck turnaround times.
Compared with static models or optimization techniques, ABS provides a more adaptable
simulation environment, especially for decentralized systems where agent interactions
significantly impact performance.
In summary, this research contributes significantly by synchronizing yard crane
scheduling strategy and truck appointment systems—both centralized and decentralized—
using agent-based simulation. It advances the field by addressing the critical gap in the
literature where previous studies either focused on one system at a time or lacked inte-
gration between yard crane and truck scheduling. This holistic approach provides a new
understanding of how to optimize terminal operations, reduce truck waiting times, improve
yard crane utilization, and reduce CO2emission, distinguishing it from prior research.
3. Agent-Based Modeling Container Pick-Up
This research used the methodology of agent-based modeling (ABM) developed by
Wilensky and Rand [
43
]. Phenomena in the container terminal, especially yard scheduling
and truck appointment, can be effectively modeled with agents, an environment (yard and
gate system), and a description of agent–agent and agent–environment interaction.
Each truck agent aims to minimize waiting time as a part of turnaround time so that
the trucks that arrive first can be served first. However, with the objective of external trucks,
the RTG transport time may increase due to activities such as back-and-forth movements
and reshuffling of containers. Therefore, the objective function of the system is to minimize
the average truck turnaround time at the container yard. With this shared objective function,
the overall container handling activities can be minimized. To achieve the common goal
while considering the independence of each agent, a utility is needed to regulate the conflict
between trucks and RTG.
3.1. Yard Crane Scheduling Strategy
When external trucks arrive in the yard, they wait for an available yard crane to service
them; crane operators must decide which truck to service first because multiple trucks
may arrive at the same block. Various sequencing methods can be used to deploy the yard
crane to the trucks, with the most common method being first-come-first-served (FCFS).
However, simulation results by Huynh and Vidal [
18
] indicated that if crane operators
select the trucks closest to them (nearest-truck-first-served, or NTFS), minimizing the need
Sustainability 2024,16, 9743 9 of 25
for cranes to turn frequently (a time-consuming process) and reverse direction, the overall
system performance improves. The NTFS approach reduces both the average waiting time
and the maximum waiting time for any truck compared with selecting trucks based on
their waiting times. For comparison, we also include the merging of time-based utility
(FCFS) and distance-based utility, nearest-truck longest-waiting-time first-served (NLFS).
In this study, we evaluate three rules: FCFS, NTFS, and NLFS. This research models the
yard scheduling strategy based on the work of Hyunh and Vidal [18] as follows.
3.1.1. First-Come-First-Served (FCFS)
Yard cranes will give higher priority to external trucks that have the longest waiting
time [18]. The utility function will calculate the waiting time of external trucks as in (1).
uc(t)FCFS = [uo(t)−ua(t)−∑2
i=1M.oi[p(c,j)]](1)
where
uc(t)
is RTG cutilization at event-time
t
.
uo(t)
is container handling service time
at event-time
t
.
ua(t)
is external truck arrival time at event-time
t
.
o1
is the presence of
another RTG on the RTG and external truck path.
o2
is RTG block change.
P(c
,
j)
is fastest
path between RTG cand target external truck j.Mis a large constant number.
3.1.2. Nearest-Truck-First-Served (NTFS)
Yard cranes will prioritize service to external trucks that are closest to the yard crane.
The utility function will calculate the distance between the yard crane and external trucks
as in (2).
uc(t)NT FS =−[T.D(c,j)+
2
∑
i=1
M.oi[p(c,j)]](2)
where
uc(t)
is RTG cutilization at event-time
t
.
T
is time required to move to the nearest
container.
D(c
,
j)
is distance between RTG cand the target external truck j.
o1
is the presence
of another RTG on the RTG and external truck path.
o2
is RTG block change.
P(c
,
j)
is
fastest path between RTG cand target external truck j.Mis a large constant number.
3.1.3. Nearest-Truck Longest-Waiting-Time First-Served (NLFS)
Vidal and Hyunh [
44
] define the time-based (FCFS) and distance-based (NTFS) into
one such as in (3).
uc(t)NL FS =−[T.D(c,j)] + uc(t)FC FS (3)
where
T
is time required to move to the nearest container.
D(c
,
j)
is distance between RTG
cand the target external truck j.uc(t)FCFS is time-based (FCFS) utility function.
3.2. Truck Appointment System Approaches
Phan and Kim [
28
] identify two approaches for adjusting truck arrival times in an
appointment system for multiple trucking companies: (1) a centralized decision-making
model (CDM) and (2) a decentralized decision-making model (DDM). CDM involves a
single decision maker, typically the container terminal, making all decisions. This model
is effective when the terminal has significant bargaining power over trucking companies.
However, because trucking companies and terminal operators are often independent
entities with their specific conditions, DDM is generally used to facilitate negotiations
between trucking companies and the terminal operator.
In this study, the truck appointment system is generally operated by port operators
who set a maximum quota of trucks allowed to enter a block during a specified period. From
the truck company’s perspective, when the quota for a particular period is reached, trucks
will arrive at another period; in this study, that process is referred to as the centralized truck
appointment system (CTAS). Another approach in truck appointment systems involves
collaboration between trucking companies and terminals to coordinate operational activities
and truck arrival schedules [
28
]. The mechanism used involves negotiation between several
Sustainability 2024,16, 9743 10 of 25
trucking companies and container terminal operators to schedule truck arrivals at the port
during peak hours; in this study, that process is referred to as the decentralized truck
appointment system (DTAS).
3.2.1. Centralized Truck Appointment System (CTAS)
CTAS is a truck arrival scheduling system wherein container terminal operators can
manage container pick-up activities based on resource availability in the field. Truck
arrivals are evenly distributed over a specific time horizon and divided into several time
windows to achieve ideal conditions. The detailed process of CTAS in this research is
described in Figure 2.
Centralized Truck Appointment System
Trucking Company Container TerminalExternal Truck
Active A gent
Propose
appointment
Assign external
truck to pick up
container
Active A gent
- Decide arrival time of
exter nal tru ck
Active Age nt
External truck
arrives at t he
container
termin al
Waiting for RTG
RTG Operator
decide truck to
be s erved
RTG Travel to
the selected
truck
Pick-up process
of container by
RTG
External truck
depart from
container
termin al
Start
Stop
Figure 2. Centralized truck appointment system (CTAS).
3.2.2. Decentralized Truck Appointment System (DTAS)
DTAS involves determining truck arrival schedules through negotiations between
container terminals and trucking companies. The decentralized negotiation model, based
on an agent-based mechanism within DTAS, considers the projected waiting times that
trucks are expected to experience within the container terminal zone. The essence of imple-
Sustainability 2024,16, 9743 11 of 25
menting the truck reservation system lies in scheduling truck arrival times. A negotiated
truck appointment system refers to a DTAS with negotiation protocols that allow trucking
companies to make changes to their arrival times at the terminal according to their preferred
truck arrival schedule. Decisions are made independently by each trucking company and
terminal. The detailed process of DTAS in this research is described in Figure 3.
Decentralized Truck Appointment System
Trucking Company Container TerminalExternal Truck
Active A gent
Propose preference
arrival time of
appointme nt
Active A gent
Activ e Agent
External truck
arrives at the
container
termi nal
Waiting for RTG
RTG Operator
decide truck to
be serv ed
RTG Travel to
the selected
truck
Pick-up pr ocess
of container by
RTG
- Calculate total waiting time for
propose ti me window
- Suggest the id eal arriva l time
External truck
depart fro m
container
termi nal
Start
Stop
Receive the
information
Receive the
appointment
Update arrival
time
Approval of truck arrival
time
Assign ext ernal
truck to pick up
containe r
Figure 3. Decentralized truck appointment system (DTAS).
4. Agent-Based Simulation Model
4.1. Conceptual Model
To develop an agent-based simulation model, NetLogo 6.3.0 software, a simulation
platform and programming language, is used [
43
]. The evaluation of integrating a yard
crane scheduling strategy and centralized/decentralized truck appointment system is
based on an agent-based simulation model that was developed in prior research [
18
,
29
].
The steps in the simulation using an agent-based model are as follows.
Sustainability 2024,16, 9743 12 of 25
1.
Determine the research questions that serve as the foundation for choosing an agent-
based model.
The research question related to the simulation being developed is as follows: how
does the external truck turnaround time change by considering the yard crane scheduling
strategy and the truck appointment system mechanism?
2. Determine the type of agent.
In this simulation, there are four types of agents: (1) external trucks, (2) containers,
(3) yard cranes
, and (4) trucking companies. The detail of each agent is described in Table 3.
Table 3. Type of agent.
Type of Agent Index Task
Container(s) n={1, 2, 3, . . . . N}Containers that arrive at the stacking yard
Crane(s) c={1, 2, 3, ...10}RTG (rubber-tired gantry) crane that serves container pick-up
Truck(s) j={1, 2, 3, . . . . N}Trucks that arrive to pick up containers
Client(s) i={1, 2, 3, . . . . N}Trucking company that assigns external truck to pick up container
3. Specify agent properties.
•
Agent status: Indicates the availability of an agent acting as a resource (available
or not). An agent with status is an agent providing port services and logistics
such as yard cranes and trucks.
•
Agent position: Indicates the geographic location of an agent placed. An agent
with a position is mobile (mobile agent) in the process, such as containers, yard
cranes, and external trucks.
•
Preference: Indicates the timing of retrieval containers. An agent with preference
is a trucking company.
•
Service time: Indicates the duration of service of the agent. An agent with service
time is a yard crane.
The details of agent properties are listed in Table 4.
Table 4. Agent properties.
Truck(s)
Variable Description
cargo Truck destination container
my-crane RTG that will or is currently serving the truck
my-utility Utility value of existing trucks
my-group Group where the destination container is located
my-stack Column where the destination container is located
my-start-time Time the truck arrives in the stacking yard
waiting True if the truck is waiting outside the port to enter because there is another truck in the same stack
on-service True if the truck is being served by the RTG
current-idle Current time spent idling (waiting to be serviced in the stack)
service-time Starting ticks when truck is being serviced
my-block Block where the destination container is located
my-terminal-time Time when the truck arrives inside the terminal
my-queue-time Total time truck spends queueing outside terminal
my-client Company that owns the truck
booked-truck Binary booking status of truck (true/false)
my-arrival-time Time when the truck booked its slot
Sustainability 2024,16, 9743 13 of 25
Table 4. Cont.
Crane(s)
Variable Description
goal Destination of an RTG (external truck that will be served)
travel-distance Distance traveled by crane
my-block Block where the destination container is located
crane-idle Number of idle cranes
crane-service Number of cranes in service
state? State of crane
t-gantry Time needed for current gantry operation
t-liftnl Time needed for current lift without load operation
t-lift Time needed for current lift with load operation
t-gantry-back Time needed for current gantry operation to the truck
gantry-position Location of the gantry in the row
Container(s)
Variable Description
z-cor Z-coordinate of the container
my-group Group where the container is located
my-stack Column where the container is located
my-row Row where the container is located
my-truck Truck that will or is currently picking up the container
my-crane RTG that will or is currently serving the container
my-block Block where the destination container is located
pick-me False if the container’s truck is not yet in the stack
Client(s)
Variable Description
my-preference Preferred arrival time
my-bound Time flexibility bound
my-ewt Expected waiting time
my-wait-time Actual waiting time for the client
my-arrival-time Final arriving time
my-truck External truck belongs to client
cargo Truck destination container
my-start-time Time the truck arrives in the stacking yard
booked-client Binary booking status of trucking company (true/false)
4. Determine the environment and stationery agent.
The agent that becomes the environment in this simulation is the main agent. This
main agent is the place where other agents are generated. Specifically, the main agent in
this simulation is the container terminal, especially the landside area where the receiving
and delivery takes place.
5. Determine agent behavior.
The behavior of each agent for the simulated study system is described along with the
time steps.
6. Designing time step.
In the design of the time steps, the behavior of each involved agent is also described.
7. Determine the parameters and performance indicators of the model.
Sustainability 2024,16, 9743 14 of 25
The performance measure in this study system is truck waiting time at the container
terminal. The desired performance measure is the minimization of average truck waiting
time at the container terminal. Truck waiting time at the container terminal affects the truck
turnaround time. Truck turnaround time also become an important indicator of container
terminals [6].
The conceptual model of the agent-based simulation is illustrated in Figure 4.
Average Truck
Turnaround Time
Truck
Turnaround
Time
Number of Serviced
Conta iner
Total External
Truck Waiting
Time
Tota l Exter nal
Truck Service
Time
Yard Cran e
Service Time
Yard Cran e
Travel Time
Yard Cran e
Transfer
Processing
Time
Yard Cran e
Lifting Time
Unit Yard C rane
Travel Time
Unit Yard C rane
Transfer Processing Time
Unit Yard C rane
Lift-On Lift-Off Time
External Truck
Waiting Time in
Container
Terminal
External Truck
Waiting Time
before Yard
Cran e Ser vice
Actual Arrival
Time
Preferred
Arrival Time
Truck
Appointment
System
Approach
Yard Cran e
Scheduling
Strategy
Yard Crane Scheduling
Conc eptu al Mo del
Truck Appointment System
Conc eptu al Mo del
Object ives
Figure 4. Conceptual model.
4.2. Agent-Based Simulation
Based on the conceptual model that has been constructed, and the method used, which
is agent-based modeling that is distributed in nature, an agent-based simulation model is
built. The simulation begins with the initialization of the global parameters, simulation
environment, container agent, crane agent, and trucking company agent. The pseudocode
of the agent-based simulation model is as follows.
Sustainability 2024,16, 9743 15 of 25
Setup:
Initialize agents, world, crane, and containers
Go:
For each session:
Initialize containers
Initialize clients
Perform appointments
Appointments are made at the beginning of each session
Reset current expected wait time
Reset current estimation at the beginning of each session
Set x to 1
Calculate new appointments count
Repeat for each new appointment:
Choose a client with booking not yet done
Set cargo’s stack
Update client’s expected wait time
Book cargo for client
If decentralized procedure is enabled:
Decentralized procedure time
Do-Arrive:
Assign clients to create their trucks if they haven’t already
Create trucks for clients who have arrived and have cargo assigned
Appointment-Arrival:
Handle appointment arrivals
Do-Move:
Count trucks threshold inside the terminal
Move the chosen truck to the container’s position
Go-Crane:
Check if it’s time to perform crane operations
If it’s time and there’s a goal position:
Move towards the goal position
Perform gantry, lift-no-load, and lift-load operations as necessary
Deliver the container if it’s at the designated position
The first stage of the agent-based simulation involves making appointments. During
the appointment process, the performance indicator to be calculated is the truck turnaround
time, which is calculated as shown in (4).
TTt=∑m
i=1∑n
j=1Tdi j −Taij
st(4)
where
•i: index for trucking company;
•j: index for container pick-up appointment by external truck;
•m: number of trucking company;
•n: number of containers pick-up;
•TTt: average truck turnaround time at event time t;
•Tdi j
: departure time for container pick-up appointment by external truck jowned by
trucking company i;
Sustainability 2024,16, 9743 16 of 25
•Taij
: arrival time for container pick-up appointment by external truck jowned by
trucking company i;
•st: number of serviced trucks at event time t.
To simulate the variability in the estimated waiting times or arrival times for each
external truck agent, the variable for individual external truck turnaround time is designed
to introduce a level of randomness that mimics real-world uncertainties, as in (5).
wijk =T Tt.RVF (5)
where
•k: index for session;
•RVF: random variation factor;
•wijk
: expected waiting time for container pick-up appointment by external truck j
owned by trucking company iat session k.
For the decentralized procedure, the container terminal will calculate the estimated
waiting time that considers both the current average truck turnaround time and the number
of appointments scheduled in the session to forecast the waiting time for the next hour, as
formulated in (6).
TTk=∑m
i=1∑n
j=1Tdi j −Taij +∑iϵk∑jϵkwi jk
st+ak
(6)
where
TTk
is the average truck turnaround time at session k.
ak
is the number of appoint-
ments at session k.
The estimated waiting time for external trucks is communicated to the trucking com-
pany in response to their appointment application, whereupon the trucking company
determines the optimal arrival time (adjusted pick-up time) based on their lower and
upper bounds, as shown in (7). This adjusted pick-up time is then submitted to the con-
tainer terminal. The appointment process concludes once the terminal operator sends an
appointment confirmation.
Tadji j =TTk−wijk +Taij ;
lbi≤Tadjij ≤ubi(7)
where
Tadji j
is the adjusted arrival time for container pick-up appointment by external
truck jowned by trucking company i.
lbi
is the earliest possible time (lower bound) of
trucking company i.ubiis the latest possible time (upper bound) of trucking company i.
To measure the performance indicator for each agent, trucking company (agent client)
and container terminal (agent crane), we propose the inconvenience cost (
ICij
) in (8) and
yard crane utilization (uc) in (9).
ICij =Tadjij −Taij (8)
uc =1−time_crane_idle
numb er o f ti cks (9)
4.3. System and Simulation Setup
The model parameterization refers to one of a container terminal in Indonesia, which
primarily uses one yard crane per block. The simulation model is adopted from the
reference models by Wasesa et al. [
29
]. The stacking yard consists of 40 container bays,
with 6 rows and a maximum height of 4 container tiers. Each time window (1 time
window:
60 min
) has 300 containers in the stacking yard awaiting the delivery process.
In modeling and setting parameters for yard crane gantry speed and handling times, this
study references Huynh and Vidal [
18
], who identify a typical yard crane gantry speed
of 135 m per minute. Consequently, it takes approximately 6 ticks for the crane to move
Sustainability 2024,16, 9743 17 of 25
from one 40-foot bay to the next. We tested four constant truck arrival rates—6, 7, 8, and 9
trucks per hour—using a single yard block and one yard crane. The details of the system
parameters are described in Table 5.
Table 5. System parameters.
Parameter Value
Number of containers 300 TEU
Number of bays 40
Number of rows 6
Number of tiers 4
Number of yard cranes 1 unit
Yard crane travel time 6 s/unit container
Yard crane transfer processing time 50 s/unit container
Yard crane lifting constant time 15 s/unit container
Yard crane lifting elevation time 5 s/unit container
Gantry movement time 1 s/unit container
Truck arrival rate per hour 6, 7, 8, 9 trucks/hour
This research evaluated alternative truck appointment system models in terms of the
performance measure of the truck turnaround time in the container terminal. In developing
the simulation model, it is essential to determine simulation inputs that correspond to
decision variables (controllable inputs). Based on the simulation scenarios, three yard crane
scheduling strategies and two truck appointment system procedures are simulated. The
inputs are represented by chooser as input strategies for yard crane scheduling and switch
as input strategies for truck appointment system procedure, as follows in Figure 5.
Figure 5. Controllable inputs for yard crane scheduling strategy (left) and truck appointment system
procedure (right).
A total of 3
×
2
×
4 scenarios are conducted, and each scenario is replicated 30 times.
The appointment process is performed at the beginning of each time window, with 10 time
windows in total, where each time window, as commonly used in container terminals, is
1 h long. The scenario designs for the simulation are described in Table 6.
Table 6. Simulation parameters.
Component Value Status
Truck appointment system mechanism Centralized, Decentralized Decision Variable
Yard scheduling strategy FCFS, NTFS, NLFS Decision Variable
Simulation duration 36,000 s Simulation Parameter
Number of time windows 10 Simulation Parameter
Duration of each time window 3600 s/time window Simulation Parameter
Warm-up duration 2 time windows (7200 s) Simulation Parameter
Replication 30 Simulation Parameter
The simulation model verification is conducted by ensuring that each entity in the
simulation model follows the process flow outlined in Figures 2and 3. Verification involves
Sustainability 2024,16, 9743 18 of 25
checking the code and logic of the simulation model. Code inspection in NetLogo is used
to confirm that no errors are present. In NetLogo, verification is performed by selecting
the “Check” option. Based on Figure 6, it is confirmed that the syntax created could be
executed, thereby verifying the model developed using NetLogo.
Figure 6. Syntax verification conducted using the “Check”’ feature.
The simulation model is validated by first testing the simulation model with the
parameters used in the reference model and comparing the results obtained. In the model,
the higher the truck arrival rate, the higher the average turnaround time for external trucks,
as observed in the reference model results. Second, the model validation is conducted
by comparing performance measures. Validation is performed for the CTAS simulation
and the FCFS yard scheduling strategy by comparing the performance metrics from the
reference study with those from the developed model, specifically focusing on the average
turnaround time. The data tested uses a truck arrival rate parameter of 6 trucks per hour.
Validation is carried out using a t-test with a 95% confidence interval. The comparison
between the reference study results and the simulation run with 30 replications shows a
significant value of 0.139, indicating that the model can be considered valid.
5. Experimental Results and Discussion
This research phase tests an agent-based truck appointment system model by com-
paring alternative configuration scenarios for the truck appointment system approach and
yard crane scheduling strategy. Tables 7and 8list the results of simulations conducted for
varying rates of external truck arrivals (in trucks per hour), along with the turnaround time
of external trucks (in seconds) for two different reservation scenarios: centralized (CTAS)
and decentralized (DTAS) procedures. Each procedure is simulated for three yard crane
scheduling strategies, FCFS, NTFS, and NFLS. For each rate of external truck arrival, there
are minimum, average, and maximum values for the turnaround time of external trucks
in both reservation scenarios. This information provides an overview of the variation in
external truck turnaround times that may occur in different situations.
In both scenarios of the truck appointment system mechanism and yard crane schedul-
ing strategy, i.e., FCFS, NTFS, NFLS, an increase in the turnaround time of external trucks
is observed to accompany an increase in the rate of external truck arrivals. This can be
seen from the significantly increased average turnaround times, with the increase in truck
arrival rates from six to nine trucks per hour.
Table 7. Results of the centralized truck appointment system (CTAS)’s turnaround time.
External Truck
Arrival Rate
(Trucks/Hour)
Turnaround Time (s)
FCFS NTFS NLFS
Min. Avg. Max. Min. Avg. Max. Min. Avg. Max.
6 247 351 498 218 254 288 241 347 474
7 308 512 730 246 275 308 357 485 357
8 679 1197 2453 262 307 354 540 1225 2458
9 1203 2380 3198 300 342 380 1011 2443 3606
Sustainability 2024,16, 9743 19 of 25
Table 8. Results of the decentralized truck appointment system (DTAS)’s turnaround time.
External Truck
Arrival Rate
(Trucks/Hour)
Turnaround Time (s)
FCFS NTFS NLFS
Min. Avg. Max. Min. Avg. Max. Min. Avg. Max.
6 251 335 439 209 248 276 274 351 488
7 339 488 724 231 275 323 352 492 783
8 659 1127 2141 269 303 331 523 1113 1882
9 1216 2301 3318 306 344 377 1519 2321 3086
At all truck arrival rate levels, the NTFS scenario consistently shows lower external
truck turnaround times than the FCFS scenario. The average turnaround times in the NTFS
scenario (254–342 s) are significantly lower than those in the FCFS scenario (351–2380 s),
indicating that NTFS is more efficient in optimizing external truck turnaround times.
The NLFS scenario presents average turnaround times ranging from 347 to 2443 s,
generally performing better than FCFS but not surpassing NTFS, especially at higher
arrival rates.
External truck turnaround times vary considerably across all three strategies and at
each arrival rate, as indicated by the difference between the minimum and maximum
times. The NTFS scenario demonstrates the smallest range of variation, suggesting it is
more consistent at providing predictable turnaround times. In contrast, the FCFS strategy
exhibits the widest range, with turnaround times varying from 247 to as high as 3198 s.
The NLFS strategy is more consistent than FCFS at lower arrival rates but shows
greater variability at higher rates, particularly with maximum turnaround times exceeding
3600 s. NTFS remains the most reliable and efficient strategy, particularly under higher
traffic conditions.
In the DTAS, each agent has the authority to make reservations, whereas in the CTAS,
truck appointments are made through a single central entity: the container terminal. De-
centralization provides greater flexibility to agents (in this case, trucking companies) to
manage reservations according to their arrival preferences. Furthermore, in the decentral-
ized approach, the main consideration is how to align truck arrival schedules as closely
as possible with trucking companies’ preferred times. This yields outcomes that are not
significantly different from those of the centralized system in terms of the performance
indicator, i.e., external truck turnaround time. The results of various configurations of truck
appointment approaches and yard scheduling strategies in terms of truck turnaround time
are described in Figure 7.
Figure 7. Comparison of truck turnaround time.
Sustainability 2024,16, 9743 20 of 25
Based on yard crane utilization, NTFS consistently shows lower utilization rates
compared with FCFS in both the centralized and decentralized truck appointment system
approaches, as presented in Tables 9and 10. This indicates that yard crane operation time
under NTFS is lower than under FCFS, aligning with the truck turnaround time results. In
other words, with NTFS, the yard crane remains available to service more external trucks
due to its more efficient operation.
Table 9. Results of the centralized truck appointment system (CTAS)’s yard crane utilization.
External Truck
Arrival Rate
(Trucks/Hour)
Yard Crane Utilization
FCFS NTFS NLFS
Min. Avg. Max. Min. Avg. Max. Min. Avg. Max.
6 69% 76% 82% 65% 70% 77% 71% 76% 83%
7 79% 88% 94% 76% 78% 83% 82% 88% 82%
8 94% 97% 99% 81% 85% 88% 94% 98% 99%
9 97% 99% 100% 86% 90% 93% 98% 99% 100%
Table 10. Results of the decentralized truck appointment system (DTAS)’s yard crane utilization.
External Truck
Arrival Rate
(Trucks/Hour)
Yard Crane Utilization
FCFS NTFS NLFS
Min. Avg. Max. Min. Avg. Max. Min. Avg. Max.
6 71% 75% 81% 64% 70% 73% 71% 76% 84%
7 80% 87% 92% 72% 78% 82% 84% 88% 93%
8 93% 97% 100% 82% 85% 90% 93% 98% 100%
9 96% 99% 100% 88% 91% 95% 96% 99% 100%
The results of various truck appointment system configurations and yard scheduling
strategies regarding yard crane utilization are shown in Figure 8.
Figure 8. Comparison of yard crane utilization.
From the data on yard crane utilization across different external truck arrival rates,
FCFS has the highest utilization, peaking at 100%, while NTFS maintains lower average
Sustainability 2024,16, 9743 21 of 25
utilization rates (70–90%), indicating better availability of yard cranes. The NLFS strategy
exhibits similar utilization trends to FCFS at lower arrival rates but approaches NTFS
utilization levels as the truck arrival rate increases, with utilization reaching up to 100% at
higher rates.
Given the significant difference in inconvenience costs, NTFS is more effective than
FCFS in minimizing inconvenience for truck drivers, leading to higher overall satisfaction
and smoother operations, especially under higher traffic conditions, as shown in Table 11
and Figure 9. Therefore, if the goal is to reduce inconvenience for truck drivers and ensure
more consistent operations, NTFS would be the superior strategy.
Table 11. Results of the decentralized truck appointment system (DTAS)’s inconvenience.
External Truck
Arrival Rate
(Trucks/Hour)
Average Inconvenience
FCFS NTFS NLFS
Min. Avg. Max. Min. Avg. Max. Min. Avg. Max.
6 10 14 25 9 11 14 11 14 20
7 15 20 32 10 12 15 15 21 37
8 23 35 69 11 13 16 23 38 72
9 35 56 94 12 15 18 32 54 82
0
10
20
30
40
50
60
70
80
90
100
6789
Inconvenience Cost
Arrival Rate per Hour (trucks/hour)
Inconvenience Cost DTAS for FCFS, NTFS, NLFS
Min-FCFS Avg-FCFS Max-FCFS
Min-NTFS Avg-NTFS Max-NTFS
Min-NLFS Avg-NLFS Max-NLFS
Figure 9. Results of inconvenience cost for decentralized truck appointment system (DTAS).
NLFS shows inconvenient levels comparable to FCFS, particularly as truck arrival
rates increase. Both NLFS and FCFS tend to exhibit higher inconvenience compared with
NTFS at medium to high arrival rates. For instance, at eight–nine trucks per hour, NLFS’s
average inconvenience (38–54) aligns closely with that of FCFS (35–56), indicating that
NLFS and FCFS perform similarly in minimizing inconvenience under heavier traffic, while
NTFS remains more effective.
Based on the experimental results of turnaround time in Tables 7and 8, the CO
2
emissions are calculated using the following parameters:
(1) Average fuel consumption of trucks: 10 liters per hour;
Sustainability 2024,16, 9743 22 of 25
(2) Carbon emission factor: 2.68 kg CO2per liter of diesel.
The results from Tables 12 and 13 demonstrate a clear comparison of CO
2
emissions
across the different yard crane scheduling strategies in both centralized (CTAS) and de-
centralized (DTAS) truck appointment systems. In Table 12, the NTFS strategy shows the
lowest average CO
2
emissions at 2193.68 kg compared with FCFS (8264.61 kg) and NLFS
(8376.10 kg). This suggests that NTFS is more effective in reducing carbon emissions in
centralized systems. Similarly, in Table 13, the NTFS strategy once again results in the
lowest average emissions at 2178.28 kg, while FCFS shows 7910.83 kg and NLFS produces
7959.01 kg. In general, DTAS shows higher efficiency in emission reduction compared with
CTAS, especially at moderate to high truck arrival rates, as evidenced by slightly lower
average emissions across strategies.
Table 12. Results of the centralized truck appointment system (CTAS)’s CO2emissions.
External Truck
Arrival Rate
(Trucks/Hour)
CO2Emissions (Gram)
FCFS NTFS NLFS
Min. Avg. Max. Min. Avg. Max. Min. Avg. Max.
6 1835.80 2615.51 3707.25 1626.22 1893.85 2146.82 1792.12 2585.79 3527.24
7 2292.63 3812.22 5431.53 1831.89 2045.67 2295.31 2654.06 3611.80 2654.06
8 5052.13 8912.16 18,262.03 1951.59 2288.27 2637.96 4022.50 9121.55 18,299.01
9 8957.88 17,718.54 23,806.94 2231.11 2546.93 2830.71 7523.08 18,185.25 26,842.66
Average 4534.61 8264.61 12,801.93 1910.20 2193.68 2477.70 3997.94 8376.10 12,830.74
Table 13. Results of the decentralized truck appointment system (DTAS)’s CO2emissions.
External Truck
Arrival Rate
(Trucks/Hour)
CO2Emissions (Gram)
FCFS NTFS NLFS
Min. Avg. Max. Min. Avg. Max. Min. Avg. Max.
6 1871.36 2497.40 3271.34 1554.87 1849.24 2057.15 2040.39 2611.73 3632.51
7 2526.79 3630.20 5390.95 1722.07 2046.79 2401.81 2622.38 3659.63 5826.99
8 4903.62 8388.79 15,939.34 2003.00 2253.00 2462.70 3892.20 8289.18 14,012.13
9 9051.80 17,126.92 24,699.44 2280.67 2564.09 2809.52 11,310.78 17,275.50 22,971.15
Average 4588.39 7910.83 12,325.27 1890.16 2178.28 2432.79 4966.44 7959.01 11,610.70
6. Conclusions and Future Work
In analyzing both CTAS and DTAS alongside the yard crane scheduling strategies of
FCFS, NTFS, and NLFS, several key insights are derived. An increase in the rate of external
truck arrivals leads to a significant rise in external truck turnaround times, a trend consistent
across all strategies. The NTFS strategy consistently produces lower average turnaround
times compared with FCFS, indicating that NTFS is more efficient in optimizing truck
turnaround times. Furthermore, the NTFS strategy demonstrates superior consistency,
with a smaller range between minimum and maximum turnaround times, suggesting
greater reliability over FCFS. In terms of CO
2
emissions, NTFS consistently achieves the
lowest average emissions across both CTAS and DTAS, as seen in Tables 12 and 13. DTAS
generally shows slightly higher efficiency in emission reduction compared with CTAS,
especially at moderate to high truck arrival rates. This makes DTAS more suitable for
scenarios where emissions reduction is a priority, as it provides better alignment with truck
company preferences by accommodating truck arrivals according to their schedules. While
DTAS offers increased flexibility, allowing each trucking company to manage reservations
according to their preferred schedules, the difference in external truck turnaround time
performance between DTAS and CTAS is minimal. However, the decentralized approach
can still enhance satisfaction from the perspective of truck companies, as it provides greater
Sustainability 2024,16, 9743 23 of 25
autonomy and flexibility than the centralized system, which is controlled by a single
central entity.
To optimize key performance indicators—such as external truck turnaround time, yard
crane utilization, and trucking company inconvenience—DTAS should be implemented in
conjunction with an effective yard scheduling strategy like NTFS. This combined approach
can better accommodate varying truck arrival rates, improve operational efficiency, and
support more sustainable terminal operations.
Future work can build upon the contributions of this research by incorporating addi-
tional yard crane scheduling strategies, such as uni-directional travel and shortest process-
ing time rule. Additionally, there is potential for further improvement of the decentralized
approach algorithm, along with exploring the impact of varying truck arrival rates un-
der different operational scenarios. Finally, incorporating real-time data integration for
more dynamic appointment scheduling could enhance the flexibility and responsiveness of
the system. These additions provide a clearer pathway for future research and potential
improvements to the current model.
Author Contributions: Conceptualization, V.N.R., A.C., R.T.C. and S.; data curation, V.N.R.; for-
mal analysis, V.N.R., A.C. and R.T.C.; funding acquisition, R.T.C.; investigation, V.N.R. and A.C.;
methodology, V.N.R., A.C., R.T.C. and S.; software, V.N.R.; supervision, A.C., R.T.C. and S.; validation,
V.N.R., A.C., R.T.C. and S.; visualization, V.N.R.; writing—original draft, V.N.R.; writing—review and
editing, V.N.R., A.C., R.T.C. and S. All authors have read and agreed to the published version of the
manuscript.
Funding: This research was funded by the Faculty of Industrial Technology, Bandung Institute
of Technology (ITB) from ITB research grant under the PPMI 2024 program No.1K/IT1.C07/SK-
KP/2024.
Data Availability Statement: The data presented in this study are available in the article.
Acknowledgments: This research would not have been possible without the guidance, wisdom, and
unwavering support of the late Senator Nur Bahagia. His dedication to academic excellence and
passion for logistics and supply chain profoundly shaped this work. During the preparation of this
work, the authors used Grammarly to improve grammar, correct typos, and provide suggestions for
alternative words or phrases. After using this tool, the authors reviewed and edited the content as
needed and took full responsibility for the content of the publication.
Conflicts of Interest: The authors declare no conflicts of interest.
References
1. UNCTAD. Review of Maritime Transport 2022; UNCTAD: Geneva, Switzerland, 2022.
2.
Caliskan, A. Seaports participation in enhancing the sustainable development goals. J. Clean. Prod. 2022,379, 134715. [CrossRef]
3.
Sciberras, L.; Silva, J.R. The UN’s 2030 Agenda for sustainable development and the maritime transport domain: The role and
challenges of IMO and its stakeholders through a grounded theory perspective. WMU J. Marit. Aff. 2018,17, 435–459. [CrossRef]
4.
Chen, G.; Govindan, K.; Golias, M.M. Reducing truck emissions at container terminals in a low carbon economy: Proposal of a
queueing-based bi-objective model for optimizing truck arrival pattern. Transp. Res. Part E Logist. Transp. Rev. 2013,55, 3–22.
[CrossRef]
5.
Huynh, N.; Walton, C.M. Robust Scheduling of Truck Arrivals at Marine Container Terminals. J. Transp. Eng. 2008,134, 347–353.
[CrossRef]
6.
Abdelmagid, A.M.; Gheith, M.S.; Eltawil, A.B. A comprehensive review of the truck appointment scheduling models and
directions for future research. Transp. Rev. 2021,42, 102–126. [CrossRef]
7.
Parola, F.; Risitano, M.; Ferretti, M.; Panetti, E. The drivers of port competitiveness: A critical review. Transp. Rev. 2016,37,
116–138. [CrossRef]
8. Notteboom, T.E. The Time Factor in Liner Shipping Services. Marit. Econ. Logist. 2006,8, 19–39. [CrossRef]
9. UN Trade and Development. Handbook of Statistics 2021; United Nations: New York, NY, USA, 2021.
10.
Zhang, X.; Zeng, Q.; Chen, W. Optimization Model for Truck Appointment in Container Terminals. Procedia Soc. Behav. Sci. 2013,
96, 1938–1947. [CrossRef]
11. Morais, P.; Ltée, R.; Lord, E. Terminal Appointment System Study; Transport Canada: Ottawa, ON, Canada, 2006.
12.
Azab, A.; Karam, A.; Eltawil, A. A simulation-based optimization approach for external trucks appointment scheduling in
container terminals. Int. J. Model. Simul. 2020,40, 321–338. [CrossRef]
Sustainability 2024,16, 9743 24 of 25
13.
Zhao, W.; Goodchild, A.V. Using the truck appointment system to improve yard efficiency in container terminals. Marit. Econ.
Logist. 2013,15, 101–119. [CrossRef]
14.
Huynn, N.; Walton, C.M.; Davis, J. Finding the number of yard cranes needed to achieve desired truck turn time at marine
container terminals. Transp. Res. Rec. 2004,1873, 99–108. [CrossRef]
15.
Boysen, N.; Emde, S. The parallel stack loading problem to minimize blockages. Eur. J. Oper. Res. 2016,249, 618–627. [CrossRef]
16.
Kim, K.H.; Lee, K.M.; Hwang, H. Sequencing delivery and receiving operations for yard cranes in port container terminals. Int. J.
Prod. Econ. 2003,84, 283–292. [CrossRef]
17.
Azab, A.; Morita, H. Coordinating truck appointments with container relocations and retrievals in container terminals under
partial appointments information. Transp. Res. Part E Logist. Transp. Rev. 2022,160, 102673. [CrossRef]
18.
Huynh, N.; Vidal, J.M. A novel methodology for modelling yard cranes at seaport terminals to support planning and real-time
decision making. Int. J. Six Sigma Compet. Adv. 2012,7, 62–91. [CrossRef]
19.
Huiyun, Y.; Xin, L.; Lixuan, X.; Xiangjun, L.; Zhihong, J.; Zhan, B. Truck appointment at container terminals: Status and
perspectives. In Proceedings of the 30th Chinese Control Decision Conference CCDC 2018, Shenyang, China, 9–11 June 2018;
pp. 1954–1960. [CrossRef]
20.
Wasesa, M.; Nijdam, P.; Muhammad, I.H.; Van Heck, E. Improving the pre-notification protocol of the containers pick-up
procedure: An agent-based approach. In Proceedings of the ICAART 2012—4th International Conference on Agents and Artificial,
Vilamoura, Portugal, 6–8 February 2012; Volume 2, pp. 190–196. [CrossRef]
21.
Chen, G.; Govindan, K.; Yang, Z. Managing truck arrivals with time windows to alleviate gate congestion at container terminals.
Int. J. Prod. Econ. 2013,141, 179–188. [CrossRef]
22.
Gracia, M.D.; González-Ramírez, R.G.; Mar-Ortiz, J.; Gracia, M.D.; González-Ramírez, R.G.; Mar-Ortiz, J. The impact of lanes
segmentation and booking levels on a container terminal gate congestion. Flex. Serv. Manuf. J. 2017,29, 403–432. [CrossRef]
23.
Huynh, N. Reducing Truck Turn Times at Marine Terminals with Appointment Scheduling. Transp. Res. Rec. J. Transp. Res. Board
2009,2100, 47–57. [CrossRef]
24.
Torkjazi, M.; Huynh, N.; Shiri, S. Truck appointment systems considering impact to drayage truck tours. Transp. Res. Part E Logist.
Transp. Rev. 2018,116, 208–228. [CrossRef]
25.
Phan, M.H.; Kim, K.H. Collaborative truck scheduling and appointments for trucking companies and container terminals. Transp.
Res. Part B Methodol. 2016,86, 37–50. [CrossRef]
26.
Riaventin, V.N.; Kim, K.H. Scheduling Appointment of Truck Arrivals at Container Terminals. Int. J. Ind. Eng. Theory Appl. Pract.
2019,25. [CrossRef]
27.
Chen, G.; Yang, Z. Optimizing time windows for managing export container arrivals at Chinese container terminals. Marit. Econ.
Logist. 2010,12, 111–126. [CrossRef]
28.
Phan, M.H.; Kim, K.H. Negotiating truck arrival times among trucking companies and a container terminal. Transp. Res. Part E
Logist. Transp. Rev. 2015,75, 132–144. [CrossRef]
29.
Wasesa, M.; Ramadhan, F.I.; Nita, A.; Belgiawan, P.F.; Mayangsari, L. Impact of overbooking reservation mechanism on container
terminal’s operational performance and greenhouse gas emissions. Asian J. Shipp. Logist. 2021,37, 140–148. [CrossRef]
30.
Islam, S.; Olsen, T.; Daud Ahmed, M. Reengineering the seaport container truck hauling process: Reducing empty slot trips for
transport capacity improvement. Bus. Process Manag. J. 2013,19, 752–782. [CrossRef]
31.
Schulte, F.; Lalla-Ruiz, E.; González-Ramírez, R.G.; Voß, S. Reducing port-related empty truck emissions: A mathematical
approach for truck appointments with collaboration. Transp. Res. Part E Logist. Transp. Rev. 2017,105, 195–212. [CrossRef]
32.
Caballini, C.; Gracia, M.D.; Mar-Ortiz, J.; Sacone, S. A combined data mining—Optimization approach to manage trucks
operations in container terminals with the use of a TAS: Application to an Italian and a Mexican port. Transp. Res. Part E Logist.
Transp. Rev. 2020,142, 102054. [CrossRef]
33.
Dekker, R.; Van Der Heide, S.; Van Asperen, E.; Ypsilantis, P. A chassis exchange terminal to reduce truck congestion at container
terminals. Flex. Serv. Manuf. J. 2013,25, 528–542. [CrossRef]
34.
Murty, K.G.; Wan, Y.-W.; Liu, J.; Tseng, M.M.; Leung, E.; Lai, K.-K.; Chiu, H.W.C. Hongkong International Terminals Gains Elastic
Capacity Using a Data-Intensive Decision-Support System. Informs J. Appl. Anal. 2005,35, 61–75. [CrossRef]
35.
Yi, S.; Scholz-Reiter, B.; Kim, T.; Kim, K.H. Scheduling appointments for container truck arrivals considering their effects on
congestion. Flex. Serv. Manuf. J. 2019,31, 730–762. [CrossRef]
36.
Guo, X.; Huang, S.Y.; Hsu, W.J.; Low, M.Y.H. Dynamic yard crane dispatching in container terminals with predicted vehicle
arrival information. Adv. Eng. Inform. 2011,25, 472–484. [CrossRef]
37.
Hwang, S.-B.; Jeong, S.-J.; Yoon, S.-W. A Study on the Heuristic-Based Yard Crane Scheduling Method Using Truck Arrival
Information. J. Korea Soc. Simul. 2019,28, 45–56. [CrossRef]
38.
Zhou, C.; Lee, B.K.; Li, H. Integrated optimization on yard crane scheduling and vehicle positioning at container yards. Transp.
Res. Part E Logist. Transp. Rev. 2020,138, 101966. [CrossRef]
39.
Gao, Y.; Ge, Y.E. Integrated scheduling of yard cranes, external trucks, and internal trucks in maritime container terminal
operation. Marit. Policy Manag. 2023,50, 629–650. [CrossRef]
40.
Talaat, A.; Gheith, M.; Eltawil, A. A Multi-Stage Approach for External Trucks and Yard Cranes Scheduling with CO
2
Emissions
Considerations in Container Terminals. Logistics 2023,7, 87. [CrossRef]
Sustainability 2024,16, 9743 25 of 25
41.
Huang, Y.; Jin, Z.; Liu, P.; Wang, W.; Zhang, D. Container drayage problem integrated with truck appointment system and
separation mode. Comput. Ind. Eng. 2024,193, 110307. [CrossRef]
42.
Ma, M.; Zhao, W.; Fan, H.; Gong, Y. Collaborative Optimization of Yard Crane Deployment and Inbound Truck Arrivals with
Vessel-Dependent Time Windows. J. Mar. Sci. Eng. 2022,10, 1650. [CrossRef]
43.
Wilensky, U.; Rand, W. An Introduction to Agent-Based Modeling: Modeling Natural, Social, and Engineered Complex Systems with
NetLogo; The MIT Press: Cambridge, MA, USA, 2015.
44.
Vidal, J.M.; Huynh, N. Building Agent-Based Models of Seaport Container Terminals. In Proceedings of the International
Conference on Autonomous Agents and Multiagent Systems (AAMAS), Toronto, ON, Canada, 11 May 2010; pp. 49–58.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.
