Investigating the Structural Dynamics of Terminal Operating System Selection: A Holistic Framework from Automation to Intelligence in Container Terminals

Abstract

In the face of mounting complexity in container terminal operations, the selection of an effective information system is paramount. The TOS (Terminal Operating System) is the most significant of all the information systems in existence for terminals. The objective of this study is to establish a set of criteria for selecting container TOS, determine the priority weights of these criteria and investigate their interactions. To the author’s knowledge, this is the first study to address this topic in such a detailed context. The hybrid FAHP (Fuzzy Analytic Hierarchy Process) and F-DEMATEL (Fuzzy Decision-Making Trial and Evaluation Laboratory) methodology was employed for the 18 criteria that were identified through the academic literature and expert views. The findings demonstrated that container terminal operators have expressed an expectation for a TOS structure that integrates complex business processes, provides effective decision support, increases traceability, works in harmony with advanced technologies, supports smart port transformation processes, enhances digital maturity and enables rapid intervention in bottlenecks. Furthermore, the fact that TOSs should support integration with external stakeholders is also critical in terms of collaboration and transparency, which are of great importance in supply chain management. It is hoped that the present study will contribute to the relevant literature and also provide a structural framework for terminal operators to select the most suitable TOS and for providers to design the most effective product.

1. Introduction

Globalisation and increasing trade volumes have led to a growing importance of maritime transport [1] and ports [2]. As the hub of logistics and the main gateway for international trade [3], ports are critical to the global economy [4]. In this context, ports are facilities that should be integrated into global supply chains [5,6] and their high performance affects both their own efficiency and competitiveness, as well as that of the supply chains [7,8]. The digitalisation and modernisation of ports is imperative [3] in such contexts, where complex logistics activities are conducted and competitive pressures are prevalent [9,10]. According to [11], the most significant factors that will influence the future of ports are sustainability, intense competition and technological developments. In this context, the impact of Industry 4.0 (which is currently affecting many sectors) on ports has been realised through the Port 4.0 (smart port) concept [12]. The concepts of Industry 4.0 and smart ports, which are based on digitalisation and autonomy, require the use of advanced technologies [13]. In this context, the increasing volume of business, the complexity of logistics processes and the significant role of ports in supply chain management [14] have brought ICT (Information and Communication Technologies), MIS (management information systems), automation and smartisation to the forefront in port management [15,16]. According to [17], the size of the relevant market is expected to reach USD 39.34 billion by 2034.

Hervás-Peralta et al. (2019) emphasised that ICT usage is an indicator of development for ports [18]. In contemporary ports, especially those with container terminals, a plethora of advanced technologies are employed [19,20], including EDI—Electronic Data Interchange [21], PCS—Port Communication System [22,23], automatic gate systems [24], AR—Augmented Reality and VR—Virtual Reality [11], digital twins [25,26,27], OCR—Optical Character Recognition [28], blockchain [29], RFID—Radio Frequency Identification [30], robots [31], AGV—Automatic Guided Vehicles [32], GPS—Global Positioning System [24,33], big data [34,35], cloud computing [33], drone [5,36], sensors [37,38], ERP—Enterprise Resource Planning [39], IoT—Internet of Things [40,41,42,43], AI—Artificial Intelligence [11,18,44,45] and TOS—Terminal Operating System [1,5]. The technologies utilised in ports and their respective domains of application are shown in Table 1.

Table 1. Emerging technologies and their use in ports.

Technologies	Areas of Use	Technologies	Areas of Use
AI	Forecasting, optimisation, security, risk management	Drone	Facility management, security, monitoring, mapping, cargo, vessel, vehicle and environmental tracking
IoT	Data collection, data sharing, security, predictive maintenance, cargo, vehicle and vessel tracking	Autonomous vehicles and robots	Occupational safety, handling, inspection, security, stacking, safety
Cloud computing	Data storage, data sharing, data processing, remote access	AR&VR	Training, 3D modelling, simulation, safety, security
Sensors	Data collection, forecasting, monitoring, tracking	Big data	Data storage and analysis, forecasting, risk management
Digital twins	Simulation, optimisation, decision support, forecasting, safety	Blockchain	Privacy, monitoring and tracking
Simulation	Traffic management, quay management, decision support, safety, scenario testing, optimisation	5G	Ultra-fast communication and data transmission
GPS	Vehicle, vessel and cargo tracking, security	OCR&RFID	Monitoring, tracking, gate and quay management, inventory management, security

Container terminals represent a pivotal element within the domain of maritime transport [46], characterised by a multitude of challenging operational processes [47]. In the context of this rapidly expanding sector, characterised by a multitude of stakeholders [48], the utilisation of ICT and MIS is important for achieving cost-effectiveness and optimal management [4]. It can be stated with ease that container terminals, which exhibit a high degree of standardisation and specialisation, are the foremost proponents of automation [49]. In this context, one of the most prominent information systems used in container terminals is TOS [19]. Inutsuka et al. (2024) defined TOS as a management information system responsible for all processes based on different components (equipment, IT systems, infrastructure) at the terminal [5]. In the contemporary context, TOSs for container terminals have undergone a transformation, evolving from supporting tools that facilitated office operations to sophisticated e-platforms [8]. It is evident that TOSs are instrumental in facilitating the seamless operations of container terminals, thereby ensuring the benefits of these systems are distributed among all stakeholders [47]. TOSs play a pivotal role in real-time planning, management, monitoring and control activities, contributing to the efficiency and effectiveness of terminal operations [50,51]. In this context, the selection of a TOS for container terminals is of great importance for effective and efficient terminal management, as well as for terminal stakeholders and, indirectly, supply chain management. As Weerasinghe et al. (2024) have observed, errors in the selection of a TOS have been demonstrated to have a detrimental effect on terminal productivity [10]. Moreover, investment and project decisions pertaining to TOSs represent a significant challenge for terminals, primarily due to the associated costs and the challenges in implementation and integration [18].

In light of this information, the study investigated the criteria to be considered in selecting a TOS for container terminals, their relative importances and the interactions between these criteria. In this context, the objective was to ascertain the answers to three research questions. The following are provided:

• Research Question (RQ) 1: In the context of current developments (technological and sectoral), what criteria should be considered in the selection of TOS for container terminals?

• RQ 2: What are the priorities of the criteria considered in TOS selection?
• RQ 3: What interactions exist between the criteria considered in TOS selection?

A set of criteria were established based on the academic literature and expert opinions. A series of face-to-face interviews were conducted with eight experts, including TOS software developers, TOS consultants, container terminal managers and academics. The findings were obtained by applying the FAHP and F-DEMATEL hybrid model to data collected from surveys administered to experts in the field. To the author’s knowledge, this study is the first to identify a set of selection criteria for container TOSs, determine the importance of these criteria and the relationships between them. It is hoped that this study will contribute to the extant literature on the subject and provide useful insights for container operators, TOS providers and stakeholders in the industry. This paper is structured as follows. Section 2 reviews the TOS including definitions, functionalities and benefits of it. Section 3 includes literature review, Section 4 reports the research methodology and provides definitions of criteria. In Section 5, the main findings are detailed. Finally, Section 6 discusses the results of analysis and summarises the main conclusions.

2. An Important Information System for Seaports: TOS

Increasing ship sizes [52], diversifying customer demands [53], shorter delivery times [54,55] and increasingly complex logistics processes [8,56], problems with real-time information sharing and integration with internal and external stakeholders [1,39,57], increasing amount of circulating cargo [58], planning and scheduling difficulties [39] have increased the need for information systems and technologies in ports (especially container ports) which are striving to integrate into global supply chains and gain a competitive advantage. In this context, cargo tracking [46], operators’ integration [59], automation [47], determining stacking areas for cargo [18], equipment synchronisation [57], optimisation and decision support [10] are of great importance in container terminals, where the TOS stands out [8,60]. Various TOS definitions in the literature are outlined in Table 2.

Table 2. Various TOS Definitions.

Reference	Definitions
[61]	Software designed to manage and control the movement of different types of cargo, in addition to supporting logistics processes.
[59]	An information system that plans, tracks and manages the movement and storage of cargo within and around the port, the use of assets and the assignment of personnel in real time.
[24]	It is an information system that integrates terminal processes and includes applications that collect, store, share, manage and analyse data, as well as aim for the efficient use of resources.
[18]	It is the main software that plans, manages and controls all processes at container terminals (e.g., handling, storage, loading, unloading, ship/truck arrival and waiting).
[62]	An integrated information system that plans, manages and controls numerous administrative and operational tasks (loading, unloading, storage, yard operations).
[3]	It is the information system that manages operations at container terminals.
[5]	The management information system responsible for all processes based on the different components (equipment, IT systems, infrastructure) at the terminal.
[63]	The core application used for planning, monitoring and executing container movements.
[46]	It is an information system used in the management and optimisation of terminal operations.

TOSs, which consist of numerous modules (e.g., vessel management, quay management, equipment management, warehousing, invoicing, reporting, gate management, damage check, intermodal operations, yard management, communication) [18,39,59], are essential for terminal efficiency, productivity, process facilitation, optimisation and automation [46]. Hervás-Peralta et al. (2019) described the TOS as the terminal’s central nervous system, on the grounds that it automates many activities at the terminal [18]. Yau et al. (2020) defined the TOS as the brain of the smart port [62], while Boer et al. (2012) defined it as the heart of ports [64]. The fundamental functions of TOSs employed in container terminals are delineated in Table 3. TOSs with the functions specified in Table 3 can provide numerous benefits for container terminals, as outlined in Figure 1. However, it should be noted that selecting an appropriate and effective TOS is crucial in realising these benefits.

Table 3. Functionalities of TOS.

Ship, quay, yard and stowage management [5,59,64,65] Management of container flow within the terminal [3] Loading/unloading planning and scheduling [39] Management of load transfers in warehouses [3] Order processing [39] Navigating external vehicles and managing traffic within the terminal [39]

Resource allocation [59] Automatic job assignment [5] Automatic billing and invoicing [8] Equipment management [59] Real-time monitoring and tracking [66,67] Support for intermodal operations [62] Reporting [51] Gate management [57]

Table 3. Functionalities of TOS.

Ship, quay, yard and stowage management [5,59,64,65] Management of container flow within the terminal [3] Loading/unloading planning and scheduling [39] Management of load transfers in warehouses [3] Order processing [39] Navigating external vehicles and managing traffic within the terminal [39]

Resource allocation [59] Automatic job assignment [5] Automatic billing and invoicing [8] Equipment management [59] Real-time monitoring and tracking [66,67] Support for intermodal operations [62] Reporting [51] Gate management [57]

Figure 1. Direct and indirect benefits of efficient TOSs. [1,3,4,5,8,11,18,23,41,46,47,57,59,60,62,64,68,69,70,71,72,73,74,75,76,77].

Figure 1. Direct and indirect benefits of efficient TOSs. [1,3,4,5,8,11,18,23,41,46,47,57,59,60,62,64,68,69,70,71,72,73,74,75,76,77].

3. Literature Review

Despite the increase in the number of studies on TOS in the literature since 2015 [78], this remains a niche topic with very few studies available. Despite the fact that the system is also applicable to a variety of cargo terminals, research on TOSs has been undertaken principally in the context of container terminals. In their study investigating the usability of ERP in conjunction with TOS, Choi et al. (2003) indicated that ERP could be particularly beneficial in data integration, quay and yard planning [39]. It has been posited that ensuring the interoperability of these two technologies could contribute to resolving the issues experienced in terminals. Gultom et al. (2024) conducted a study to ascertain the impact of the TOS utilised at the Tanjung Priok port in Indonesia on handling efficiency [3]. This investigation employed qualitative analysis as the primary methodological approach. The findings demonstrated that efficiency had increased by 116 per cent, while the ship waiting times in port had decreased. Eichenhofer et al. (2020) conducted an examination of data flows occurring via the TOS at container terminals and provided an assessment of potential security vulnerabilities [4]. The findings confirmed the existence of security vulnerabilities and revealed that TOS performance affects cargo security and on-time delivery. Min et al. (2017) conducted an examination of operational practices at a port in Korea [59]. The study’s findings led to the proposal of a TOS that was deemed to have the potential to enhance terminal efficiency. The present study emphasised the importance of terminal operators integrating with the system and compared the efficiency of two operators, one of which had integrated with the system and one which had not. Furthermore, the critical success factors of terminal operators were evaluated using the AHP method. The study posited a TOS comprising six modules (quay, ship, resources, site, communication and interface), with the operator’s cargo attraction identified as the paramount success factor. Singgih et al. (2016) emphasised the importance of synchronisation and data transfer between equipment used in automated container terminals and proposed a TOS that provides real-time decision support and reduces the berthing times of incoming ships [57]. In their study on the utilisation of information technologies in smart ports, Yau et al. (2020) categorised the features that TOSs should possess into three distinct categories [62]. These are defined as autonomous decision-making, the ability to work with different devices and equipment and real-time information sharing. Furthermore, emphasis was placed on the significance of features such as stability, scalability, reliability, accuracy and availability. Gekara & Nguyen (2020) conducted an examination of the issues experienced with the TOS application that had been implemented at the Port of Mombasa [8]. The findings indicated that the challenges experienced in the adoption and implementation of a TOS stemmed from deficiencies in ICT infrastructure and a lack of employee preparedness. Inutsuka et al. (2024) conducted a comparative analysis of the TOSs (CTMS and OPUS Terminal) utilised in Japan, along with the challenges encountered [5]. The study put forward a series of recommendations for the enhancement of the smart port concept. In their evaluations, the researchers identified six functions that should be present in TOSs in the context of the smart port concept. These are communication with the equipment control system; AI and digital twins-supported operational planning and control; mobile terminal-based work order instruction; information exchange with stakeholders via PCS; smart gate management and IoT-based data collection and sharing with other systems. Boer & Saanen (2012) identified a testing and tuning method that could positively impact the increased functionality of TOSs and the reduction in interruptions [64]. Yi et al. (2008) proposed a TOS consisting of seven main modules (basic data, entry and exit gates, internal transporting, warehousing, unloading and loading) in the context of container terminal business processes and emphasised the most important functions of the TOS as equipment control, handling, ship stowage and yard planning [47]. El Nakib & Ismail (2025) conducted a comprehensive examination of the TOS implemented at Alexandria Port, employing SWOT analysis to assess its effectiveness [46]. The study yielded actionable recommendations for enhancing the TOS’s efficacy. A comparison was also made between the efficiency levels of the pre- and post-TOS periods, with the Data Envelopment Analysis (DEA) method. The findings confirmed that the use of TOS increased efficiency. Hervás-Peralta et al. (2019) identified the critical functions that TOSs should possess in the context of seaports and dry ports [18]. The 107 identified functions were grouped into six main categories: warehouse, ERP control panel, master data, maritime operations, communication and gate. The findings from their analysis, which employed the AHP method, indicated that the most significant functions were “ship time tracking”, “development of loading/unloading lists” and “optimisation of space and cargo locations”. Sahraoui et al. (2023) investigated the impact of TOS on port operations based on KPIs (collaboration, efficiency, productivity, reputation, reliability) specifically at the Port of Rhodes in Tunisia [1]. In this particular context, the results obtained from the surveys revealed that operational efficiency was higher after TOS implementation, but there were problems with adaptation.

4. Research Methodology

To the author’s knowledge, no study has been found in the literature where the set of selection criteria for a container TOS has been determined, weighted and the interactions between the criteria analysed. The establishment of the set of criteria is predicated on a comprehensive review of studies on the selection of different information technologies and the use of the TOS in all terminal types, as well as a synthesis of expert opinions. The framework of the research methodology is illustrated in Figure 2.

Figure 2. Framework of research methodology.

4.1. Determination and Definitions of Criteria

The selection of experts for the TOS selection process at the container terminal was based on two fundamental criteria: direct decision-making authority and familiarity with the process. This enhancement in the methodology’s applicability and reliability is further substantiated by [79]. In-person interviews were conducted with pertinent individuals during this process. Given that the process is expected to be lengthy and that meetings will be held frequently, specialists in accessible and nearby locations (İstanbul and Kocaeli) have been preferred. Following extensive consultation with the relevant experts, a consensus was reached on all criteria. In this study, the FAHP method was employed to assign weights to the criteria, while the F-DEMATEL method was used to determine the interactions between the criteria. The surveys for both methodologies were administered to the same experts and conducted face-to-face. The qualifications of the experts are shown in Table 4. The interviews were conducted in 2025. The 18 criteria that were determined for the study, along with the relevant details, are shown in Table 5.

Table 4. Qualifications of the experts.

Experts	Experiences	Job Title	Education Level
1	12 years	Port manager	MS in Business
2	7 years	Port IT manager	MS in Computer Engineering
3	8 years	TOS developer	MS in Computer Engineering
4	18 years	Container operations manager	MS in Business
5	11 years	Port IT manager	MS in Computer Engineering
6	14,5 years	Container operations manager	MS in Mechanical Engineering
7	15 years	Academician	PhD in Port Management
8	21 years	TOS consultant/Academician	PhD in Maritime Transportation Management Engineering

Table 5. TOS selection criteria for container terminals.

Criteria	Definitions	References
Customisation (C1)	It refers to the extent to which the TOS can be customised to meet the terminal’s unique operational requirements without costly re-engineering or supplier dependency.	[8,49]
Stability (C2)	It refers to the TOS’s ability to perform consistently and correctly under defined conditions over time with minimal downtime or failures.	[62,64,80]
Functionality (C3)	It refers to the scope and structure of the features, tasks and modules (e.g., berth planning, vessel planning, yard planning, intermodal operations, billing, KPI reporting) which can provide support for the execution of terminal operations.	[1,4,64,80]
Provider Support (C4)	It refers to the level and reliability of the technical and support services provided by the TOS provider to its customers. It is important that issues pertaining to high-cost and strategic information technologies, such as TOS, are addressed promptly to ensure operational efficiency and maintain optimal functionality.	[4,80]
Root Cause Analysis (C5)	Root cause analysis is defined as a systematic process that aims to identify the fundamental causes of problems through analysis. In this context, it is important to ascertain the underlying causes of inefficiencies and planning errors that have been observed at the terminal. This criterion is indicative of the TOS’s capacity to conduct analyses based on the collected data.	Expert views
Environmental Sustainability Tracking (C6)	It refers to the TOS’s ability to measure, monitor, analyse and report on the environmental impacts of activities carried out at the terminal and constitutes the fundamental objective of this initiative. In this context, the objective is to achieve sustainability goals by monitoring environmental indicators such as carbon footprint, energy consumption, noise pollution, waste management, fuel usage and emission rates in real time or periodically.	Expert views, [62,81]
Total Cost of Ownership (C7)	It refers to all financial expenditures incurred in the process of procuring, implementing, operating, upgrading and maintaining the TOS.	[80,82]
User-Friendliness (C8)	It refers to the ease with which the TOS interface and workflows can be learnt and their effectiveness for various terminal users, from gate operators to planners.	[8,83]
Maintainability (C9)	It refers to the ease with which updates to the TOS can be made, repaired, configured or improved over time without significant disruption to operations.	[84,85]
External Integration Capability (C10)	It refers to the TOS’s ability to continuously and rapidly exchange basic data related to port and terminal operations between the terminal and the systems of various stakeholders (e.g., shipping companies, carriers, customs, ship owners and agents) via EDI and similar systems. This approach ensures that all relevant parties have access to accurate and up-to-date information, enabling them to make informed decisions, coordinate actions and optimise the terminal’s overall efficiency. It has been demonstrated that this can also contribute to the effective planning of processes, the rapid response to customers and the management of supply chains. In this context, it is important that the TOS can function seamlessly with external systems such as EDI and customs systems.	Expert views, [1,5,18,62]
AI powered optimisation and decision support (C11)	Container terminals carry out many complex activities, resulting in the collection of vast amounts of data. The use of artificial intelligence and similar technologies is essential to analyse this data and derive meaningful results (e.g., for yard planning and optimisation, equipment scheduling, berth planning). In this context, in line with current trends in information technology, TOSs are undergoing a transition towards big data analytics and artificial intelligence-based functions. This criteria refers to TOS’s capacity to enhance operational efficiency, planning and the decision-making process through the utilisation of artificial intelligence (AI) technologies.	Expert views, [18,24,57,62]
Deployment Type (C12)	It refers to the method and environment (e.g., on-premise, cloud-based) in which TOS is installed. This is of particular significance in terms of ease of access and reliability for users.	Expert views, [24]
Scalability (C13)	It refers to the TOS’s ability to manage elevated demand, enhanced complexity and augmented operational scale without any diminution in performance or the necessity for architectural modifications.	[62,85]
Simulation Capability (C14)	The modelling and testing of processes at the terminal in advance will be beneficial in terms of demonstrating the effects of new equipment and what might happen in unexpected situations. It refers to TOS’s capacity to replicate terminal operations in a virtual environment (e.g., digital twins) enables the testing of diverse scenarios, the analysis of resource planning, the anticipation of bottlenecks, the generation of reports and the visualisation of the potential impact of operational decisions.	Expert views, [86,87]
Internal Integration and automation capability (C15)	The foremost requirement for terminal automation is the interoperability of the terminal’s internal systems, equipment and technologies, which are interconnected to ensure smooth and efficient operations. Achievement of these aims is dependent on effective real-time data collection and sharing. TOSs have the capacity to operate in conjunction with a wide range of technologies (e.g., RFID for container tracking, IoT and sensors for data collection) and equipment (e.g., ASC, AGV, QC, RTG, STS). This refers to the TOS’s ability to seamlessly connect, communicate and exchange data with internal systems. In this particular context, it is imperative for the TOS to possess robust integration capabilities to achieve a high level of automation at the container terminals. The successful automation of processes, including gate management, container stacking or job assignments relies on these criteria.	Expert views, [1,4,59,62,64,88]
Real time monitoring (C16)	It refers to the TOS’s ability to continuously monitor, display and update the status of terminal operations, equipment and cargo without delay.	Expert views, [1,40,67]
Cyber Security Posture (C17)	New generation ports are becoming digitalised and integrated with numerous information systems. This situation gives rise to a number of significant concerns with regard to security and privacy. In the period between 2017 and 2022, there was a 900% increase in cyberattacks on ports. These attacks have been shown to result in operational disruptions, accidents and loss of data and money at terminals. In this context, it refers to the ability of the TOS to protect the terminal’s data, systems and operations against unauthorised access, cyber threats and data breaches.	Expert views, [4,64,89,90,91]
Compliance With Regulatory Frameworks (C18)	This refers to the ability of the TOS to operate in compliance with all relevant laws, regulations and green port concept standards (e.g., ISPS, MARPOL, ISM Code) within the scope of all terminal operations and to facilitate the execution of legal, safe and secure operations.	Expert views, [1,85,92]

4.2. Fuzzy AHP Methodology

AHP (Analytical Hierarchy Process) is a multi-criteria decision-making method that was introduced to the academic literature by Saaty in 1977. It is characterised by a hierarchical and subjective structure and is based on pairwise comparisons of criteria. This approach, utilised in the context of both criterion weighting and the evaluation of alternatives, employs Saaty’s ((1)–(9)) rating scale [93]. The FAHP method is a hybrid approach that integrates the principals of the AHP method with fuzzy logic concepts. The objective of this integration is to minimise uncertainty and subjectivity in the decision-making process. Triangular fuzzy numbers (Equation (1)) are a common mathematical construct employed in this methodology [94].

$$\widetilde{a_{ij}}=\left(l_{ij}, m_{ij}, u_{ij}\right) where \left\{\begin{array}{c}{l}_{ij}=lower bound\\ m_{ij}=most likely value\\ u_{ij}=upper bound\end{array}\right.$$

(1)

One of the applications developed is Buckley’s approach, introduced to the literature in 1985, which uses the geometric mean method in a fuzzy environment. This method, utilised in numerous studies within the extant literature [95,96,97,98,99], comprises the following steps.

In the initial phase of the method, a hierarchical structure is established, comprising objectives and criteria, sub-criteria (if applicable) and alternatives (if applicable). In the subsequent stage of the process, experts are tasked with the generation of pairwise comparisons between criteria, utilising the linguistic terms shown in Table 6 [93].

Table 6. Fuzzy AHP scale of relative importance.

Linguistic Term	Scalar Value (AHP)	Fuzzy Value (FAHP)	Inverse Fuzzy Number
Equal	1	(1,1,1)	(1,1,1)
Equal Importance	1	(1,1,3)	(1/3,1,1)
Moderate Importance	3	(1,3,5)	(1/5,1/3,1)
Strong Importance	5	(3,5,7)	(1/7,1/5,1/3)
Very Strong Importance	7	(5,7,9)	(1/9,1/7,1/5)
Extreme Importance	9	(7,9,9)	(1/9,1/9,1/7)

These comparisons are then formalised within pairwise comparison matrices, which are created for each individual decision maker. Equation (2) presents the pairwise comparison matrix ($\widetilde{C_i}$), which belongs to decision maker $i$.

$${\tilde{C}}_i=\left[\begin{array}{cccc}\left(\mathrm{1,1,1}\right)& {\tilde{a}}_{12}& \dots & {\tilde{a}}_{1n}\\ {\displaystyle {\tilde{a}}_{21}}& \left(\mathrm{1,1,1}\right)& \dots & {\tilde{a}}_{2n}\\ ⋮& ⋮& \dots & ⋮\\ {\tilde{a}}_{n1}& {\tilde{a}}_{n2}& \dots & \left(\mathrm{1,1,1}\right)\end{array}\right]\mathit{where} {\tilde{a}}_{ij}=(l_{ij},m_{ij},u_{ij})$$

(2)

In the subsequent stage, the values constituting the aggregated group judgement matrix are obtained using geometric mean, with the consideration of the pairwise comparison matrices of all experts ($K$ experts).

$$l_{ij}={\left({\prod }_{k=1}^K{l}_{ij}^{\left(k\right)}\right)}^{1/K}, m_{ij}={\left({\prod }_{k=1}^K{m}_{ij}^{\left(k\right)}\right)}^{1/K}, u_{ij}={\left({\prod }_{k=1}^K{u}_{ij}^{\left(k\right)}\right)}^{1/K}$$

(3)

Subsequently, the fuzzy geometric means are calculated for each criterion using Equation (4). The sum of the found values is taken and the weight values based on the criteria are calculated using Equation (5) [93].

$$\widetilde{r_i}={\left({\prod }_{j=1}^n\widetilde{C_{ij}}\right)}^{1/n}, i=1,2,\dots ,n$$

(4)

$$\widetilde{w_i}=\widetilde{r_i}\otimes {({\sum }_{i=1}^n\widetilde{r_i})}^{-1}=({lw}_i,{mw}_i,{uw}_i)$$

(5)

In the subsequent step, the total vector of each $\widetilde{r_i}$ is found, its (−1)st power is taken and it is multiplied by each $\widetilde{r_i}\prime$s inverse vector to ascertain the fuzzy weight ($\widetilde{w_i}$) of the i-th criterion. In this context, given that the values ($\widetilde{w_i}$) remain fuzzy values, the defuzzification process is conducted in accordance with Equation (6). In the final stage of the process, the values obtained through utilisation of Equation (6) are normalised through application of Equation (7) [98,100,101].

$$M_i={\displaystyle \frac{(u_{wi}-l_{wi})}{3}}+{\displaystyle \frac{(m_{wi}-l_{wi})}{3}}+l_{wi}={\displaystyle \frac{{lw}_i+{mw}_i+{uw}_i}{3}}$$

(6)

$$w_i={\displaystyle \frac{M_i}{{\sum }_{i=1}^n{M}_i}}$$

(7)

In order to calculate the consistency of the comparison matrices of expert judgments, it is necessary to undertake the process of defuzzification of the elements contained within the fuzzy decision matrices. This will result in the creation of a crisp pairwise comparison matrix. For the defuzzification process, the median value with the highest membership degree among the fuzzy numbers in the pairwise comparison matrix is adopted and in accordance with Saaty’s proposal, the consistency checks are to be implemented. The calculation of criterion weights can then proceed, provided that consistency is ensured. In this context, Saaty (1980) asserts that results are deemed reliable if the consistency ratio (CR) is less than 0.1 [102]. The following equations are employed to verify the consistency ratio:

$$CR={\displaystyle \frac{CI}{RI}}$$

(8)

$$CI={\displaystyle \frac{1}{n(n-1)}}{\sum }_{i=1}^n{\sum }_{j=1}^n(a_{ij}{\displaystyle \frac{w_j}{w_i}}-1)$$

(9)

where $CI$ means consistency index and $RI$ shows random index [101].

4.3. Fuzzy DEMATEL

The Fuzzy DEMATEL method, which emerged from the combination of the DEMATEL method introduced to the literature by Gabus and Fontela with fuzzy set theory, can be used to analyse cause–effect relationships between criteria and to weight criteria [103]. The method under discussion has been employed in numerous studies [103,104,105,106,107,108,109] and consists of the following steps.

Subsequent to the determination of the criteria and the experts to be evaluated, the experts are to evaluate the criteria in pairs. At this stage, the expert evaluations are converted into triangular fuzzy numbers according to Table 7 and set up direct relation matrix for each expert [106,110].

Table 7. Triangular fuzzy numbers of expert linguistic variables.

Linguistic Variable	Influence Score	Triangular Fuzzy Numbers
No influence	0	(0,0,0.25)
Very low influence	1	(0,0.25,0.5)
Low influence	2	(0.25,0.5,0.75)
High influence	3	(0.5,0.75,1)
Very high influence	4	(0.75,11)

Following this stage, the direct-relation-matrixes should be normalised (see Equations (10)–(13)) [107].

$${∆}_{min}^{max}=max{r}_{ij}^n-min{l}_{ij}^n$$

(10)

$${xr}_{ij}^n={\displaystyle \frac{r_{ij}^n-min{l}_{ij}^n}{{∆}_{min}^{max}}}$$

(11)

$${xm}_{ij}^n={\displaystyle \frac{m_{ij}^n-min{l}_{ij}^n}{{∆}_{min}^{max}}}$$

(12)

$${xl}_{ij}^n={\displaystyle \frac{(l_{ij}^n-min{l}_{ij}^n)}{{∆}_{min}^{max}}}$$

(13)

Subsequent to these operations, calculation of the right (${xrs}_{ij}^n$) and left (${xls}_{ij}^n$) normalised values is conducted in accordance with Equations (14) and (15).

$${xrs}_{ij}^n={\displaystyle \frac{{xr}_{ij}^n}{(1+{xr}_{ij}^n-{xm}_{ij}^n)}}$$

(14)

$${xls}_{ij}^n={\displaystyle \frac{{xm}_{ij}^n}{(1+{xm}_{ij}^n-{xl}_{ij}^n)}}$$

(15)

In the following step it is obtained crisp values (${\mathrm{x}}_{\mathrm{i}\mathrm{j}}^{\mathrm{n}}$) by using Equation (16) and then it is created total normalized crisp values (${\omega }_{ij}^k$) by Equation (17).

$$x_{ij}^n=\left[{\displaystyle \frac{{xls}_{ij}^n \left(1-{xls}_{ij}^n\right)+{xrs}_{ij}^n\ast {xrs}_{ij}^n}{(1-{xls}_{ij}^n+{xrs}_{ij}^n)}}\right]$$

(16)

$${\omega }_{ij}^k=\mathit{min}{l}_{ij}^n+x_{ij}^n{∆}_{min}^{max}$$

(17)

Subsequent to the execution of all operations, the normalised crisp values obtained from all experts are integrated to obtain the direct relation matrix ($M$) as in Equation (18). Furthermore, the normalised direct matrix ($X$) is derived using Equations (19) and (20) [106].

$$M=\left[{\sum }_{k=1}^E{\omega }_{ij}^k\right]/E$$

(18)

$$P={\displaystyle \frac{1}{{max}_{1\le i\le n}{\sum }_{j=1}^n{x}_{ij}}}\hspace{1em},\hspace{1em}i,j=\mathrm{1,2},\dots n$$

(19)

$$X=P\ast M$$

(20)

After the identity matrix (${\left[I\right]}_{nxn}$) is created, the normalised direct matrix ($X$) is substracted from this matrix. The obtained matrix is inversed ((${I-X)}^{-1})$. The final operation of this stage is the multiplication of the matrix ((${I-X)}^{-1}$) and normalised direct matrix ($O$) as Equation (21). This operation yields the total relation matrix ($T$).

$$T=X({I-X)}^{-1}$$

(21)

The final stage of the methodology is the calculation of $R_i$ and $C_j$ values. It can be demonstrated that $R_i$ is equivalent to the sum of the rows of the $T$ matrix, whilst $C_j$ is the sum of its columns. In addition, $(R_i+C_j)$ is the importance and $(R_i-C_j)$ is the net effect of criteria $i$. In this context, a high value of $(R_i+C_j)$ signifies that the criterion possesses significant importance within the system, while a positive value of $(R_i-C_j)$ indicates that the criterion functions as a cause and a negative value denotes that the criterion is a result criterion. In the context of drawing causal diagrams, the terms $R_i$ and $C_j$ are utilised to denote specific values. $(R_i+C_j)$ is horizontal, while $(R_i-C_j)$ forms the vertical axis. The horizontal axis of the graph represents the importance degrees of the criteria, while the vertical axis shows the effect of the relationship. In the event of the $(R_i-C_j)$ axis being positive, the factor is deemed to be in the cause group; conversely, the factor is classified as being in the effect group. Furthermore, it is imperative to calculate the threshold value from the $T$ matrix in order to facilitate the process of diagrammatic representation. To this end, the mean of the values comprising the matrix is calculated [106,110].

5. Case Study

5.1. FAHP Application and Findings

The consistency ratios for the pairwise comparison matrices created based on the assessments of eight experts within the context of the FAHP methodology are shown in Table 8. The RI value was accepted as 1.615, as there were 18 criteria [111]. The Excel programme was utilised for the execution of all subsequent calculations.

Table 8. Consistency ratios of expert judgements.

Experts	${\mathit{\lambda }}_{\mathit{m}\mathit{a}\mathit{x}}$	$\mathit{C}\mathit{I}$	$\mathit{C}\mathit{R}$ ($\mathit{C}\mathit{I}$/$\mathit{R}\mathit{I}$)	Consistency
1	19.52	0.089692094	0.056	√
2	20.14	0.125956834	0.078	√
3	20.32	0.136501907	0.085	√
4	20.46	0.144746011	0.090	√
5	20.72	0.159884249	0.099	√
6	20.58	0.151530672	0.094	√
7	20.70	0.158662386	0.098	√
8	20.27	0.133711132	0.083	√

In the subsequent stage of the method, the aggregated fuzzy judgement matrix shown in Table 9 was obtained by using Equation (3) on the values of the pairwise comparison matrices formed by expert evaluations.

Table 9. Aggregated fuzzy judgement matrix.

Criteria	C1			C2			…			…			C18
	l	m	u	l	m	u	l	m	u	l	m	u	l	m	u
C1	1.00	1.00	1.00	0.15	0.21	0.42	…	…	…	…	…	…	0.17	0.26	0.58
C2	2.41	4.68	6.78	1.00	1.00	1.00	…	…	…	…	…	…	1.00	1.51	3.63
C3	1.72	3.34	5.32	0.51	0.76	1.15	…	…	…	…	…	…	1.00	1.15	2.79
C4	0.58	0.87	2.12	0.18	0.30	0.58	…	…	…	…	…	…	0.16	0.24	0.50
C5	0.82	1.22	2.35	0.16	0.24	0.50	…	…	…	…	…	…	0.19	0.33	0.54
C6	0.67	1.07	2.05	0.17	0.26	0.58	…	…	…	…	…	…	0.17	0.28	0.54
C7	2.26	4.49	6.57	0.39	0.76	1.00	…	…	…	…	…	…	0.58	0.87	1.61
C8	1.00	1.32	2.59	0.18	0.28	0.66	…	…	…	…	…	…	0.36	0.62	1.11
C9	0.18	0.28	0.71	0.11	0.14	0.19	…	…	…	…	…	…	0.15	0.21	0.39
C10	1.50	2.62	4.35	0.21	0.39	0.76	…	…	…	…	…	…	0.41	0.87	1.32
C11	2.10	4.08	5.54	0.24	0.50	1.00	…	…	…	…	…	…	0.39	0.76	1.15
C12	0.21	0.33	0.62	0.12	0.15	0.22	…	…	…	…	…	…	0.15	0.22	0.44
C13	0.48	0.87	1.72	0.17	0.28	0.54	…	…	…	…	…	…	0.18	0.30	0.62
C14	1.22	2.70	4.31	0.26	0.58	1.00	…	…	…	…	…	…	0.39	0.74	0.87
C15	2.26	4.49	6.57	0.39	0.76	1.00	…	…	…	…	…	…	1.00	1.28	2.12
C16	2.10	4.39	6.50	0.32	0.66	1.00	…	…	…	…	…	…	0.68	0.82	1.15
C17	0.16	0.26	0.44	0.14	0.19	0.32	…	…	…	…	…	…	0.14	0.20	0.34
C18	1.73	3.87	5.92	0.28	0.66	1.00	…	…	…	…	…	…	1.00	1.00	1.00

As illustrated in Table 10, the fuzzy geometric mean values have been calculated for each criterion using Equation (4).

Table 10. Fuzzy geometric means of criteria.

Criteria	l	m	u
C1	0.371363146	0.624962592	1.074338473
C2	1.573027297	2.637856878	4.334011164
C3	1.503447669	2.497652391	3.982216004
C4	0.384067504	0.610460449	1.168688989
C5	0.437637418	0.720419933	1.263012512
C6	0.389786559	0.674187072	1.170270911
C7	1.326680472	2.370297291	3.606314055
C8	0.502002093	0.795828834	1.545681053
C9	0.187764063	0.281895805	0.508175922
C10	0.73533671	1.379897117	2.242424794
C11	0.944000579	1.802049821	2.740893123
C12	0.192332917	0.281705213	0.523009412
C13	0.309113324	0.527221328	0.924099161
C14	0.806903562	1.617213882	2.396746329
C15	1.449416541	2.271718495	3.338399008
C16	1.171881922	1.960967457	3.026673316
C17	0.207709175	0.296406365	0.54443398
C18	1.060454302	1.844964463	2.801036915
1/Sum of rows	0.026888641	0.043111429	0.073784809

In the subsequent stage, the column totals of the values displayed in Table 10 were calculated and fuzzy weight values for each criterion were obtained using Equation (5). The values under consideration are triangular fuzzy numbers and the defuzzified values ($M_i$) should be calculated using Equation (6).The normalisation of the defuzzified values was calculated using Equation (7) and the weight values ($w_i$) for each criterion were obtained. The values obtained from these analyses are collectively displayed in Table 11. The overall CR of the expert group’s judgement is 0.0249.

Table 11. Defuzzification values and weights of criteria.

Weights	${\mathit{l}\mathit{w}}_{\mathit{i}}$	${\mathit{m}\mathit{w}}_{\mathit{i}}$	${\mathit{u}\mathit{w}}_{\mathit{i}}$	${\mathit{M}}_{\mathit{i}}$	${\mathit{w}}_{\mathit{i}}$
w1	0.00998545	0.02694303	0.079269859	0.03873278	0.03
w2	0.042296567	0.113721779	0.319784186	0.158600844	0.12
w3	0.040425665	0.107677363	0.293827047	0.147310025	0.11
w4	0.010327053	0.026317822	0.086231494	0.04095879	0.03
w5	0.011767476	0.031058333	0.093191137	0.045338982	0.03
w6	0.010480831	0.029065168	0.086348216	0.041964738	0.03
w7	0.035672635	0.102186903	0.266091193	0.134650244	0.10
w8	0.013498154	0.034309318	0.114047781	0.053951751	0.04
w9	0.005048721	0.012152931	0.037495663	0.018232438	0.01
w10	0.019772205	0.059489336	0.165456885	0.081572809	0.06
w11	0.025382893	0.077688942	0.202236275	0.10176937	0.07
w12	0.005171571	0.012144714	0.03859015	0.018635478	0.01
w13	0.008311637	0.022729265	0.06818448	0.033075127	0.02
w14	0.02169654	0.069720401	0.17684347	0.089420137	0.07
w15	0.038972841	0.09793703	0.246323133	0.127744335	0.09
w16	0.031510313	0.084540109	0.223322512	0.113124311	0.08
w17	0.005585018	0.012778502	0.040170957	0.019511492	0.01
w18	0.028514175	0.079539054	0.206673973	0.104909068	0.08

The weightings ($w_i$) of the criteria are presented in Figure 3, with the ranking from highest to lowest. Accordingly, the top five most important criteria are, in order: stability, functionality, TCO, internal integration and automation capability and real-time monitoring. The criterion of maintainability is of least importance. In this context, the uninterrupted operation of the TOS, its error-free performance, consistent performance and its ability to meet functional requirements are of great importance. This situation is consistent with the complexity of processes at container terminals. It is evident that TOSs are costly and investment decisions are contingent on terminal budgets. The criteria of internal integration and automation capability and real-time monitoring, which follow in order of importance, demonstrate the high need for integration and automation of terminal business processes. In this particular context, it is evident that container terminal operators assign significant importance to concepts such as operational efficiency, process integration and real-time control. This is closely related to the need for close monitoring of container terminal operations, rapid intervention and transparency. Furthermore, it is imperative to adhere to the regulatory frameworks that govern container terminals. From this standpoint, it is important that the TOS functions in accordance with these stipulations. Furthermore, the high ranking of optimisation and simulation capability criteria demonstrates the importance that container terminals place on becoming smarter. The subsequent two most significant criteria, following user-friendliness, are root cause analysis and environmental sustainability tracking criteria. It is important to note that these criteria take precedence over criteria such as provider support and customisation. The initial situation indicates that terminal operators encounter difficulties in identifying the underlying causes of certain problems experienced at terminals. The latter situation demonstrates that terminal operators have a growing awareness of environmental sustainability and wish to achieve this goal through technological and digitalisation solutions facilitated by the TOS.

Figure 3. Graphical representation of criterion weights.

5.2. F-DEMATEL Application and Findings

In the second part of the application phase of the study, interactions between criteria were examined. The F-DEMATEL method was utilised for this purpose. To this end, F-DEMATEL questionnaires were also administered to the experts consulted for the FAHP application. In view of the fact that there were eight experts and 18 criteria, the results obtained by applying the steps up to the creation of the direct relation matrix are shown only for Expert 1. The results are presented in Table 12.

Table 12. Corresponding triangular fuzzy values for Expert 1.

Criteria	C1	C2	C3	C4	…	C18
C1	(0,0,0.25)	(0.5,0.75,1)	(0.25,0.5,0.75)	(0.5,0.75,1)	…	(0,25,0,5,0,75)
C2	(0.5,0.75,1)	(0,0,0.25)	(0.5,0.75,1)	(0.25,0.5,0.75)	…	(0,25,0,5,0,75)
C3	(0.25,0.5,0.75)	(0.5,0.75,1)	(0,0,0.25)	(0.25,0.5,0.75)	…	(0,5,0,75,1)
C4	(0.25,0.5,0.75)	(0.5,0.75,1)	(0.5,0.75,1)	(0,0,0.25)	…	(0,5,0,75,1)
C5	(0.25,0.5,0.75)	(0.25,0.5,0.75)	(0.5,0.75,1)	(0.25,0.5,0.75)	…	(0,25,0,5,0,75)
C6	(0,0.25,0.5)	(0.25,0.5,0.75)	(0.25,0.5,0.75)	(0,0.25,0.5)	…	(0,0,25,0,5)
C7	(0.25,0.5,0.75)	(0.25,0.5,0.75)	(0,0.25,0.5)	(0.25,0.5,0.75)	…	(0,25,0,5,0,75)
C8	(0.25,0.5,0.75)	(0,0.25,0.5)	(0.25,0.5,0.75)	(0.25,0.5,0.75)	…	(0,0,25,0,5)
C9	(0.25,0.5,0.75)	(0.25,0.5,0.75)	(0.25,0.5,0.75)	(0,0.25,0.5)	…	(0,5,0,75,1)
C10	(0.25,0.5,0.75)	(0.5,0.75,1)	(0.7,1,1)	(0,0.25,0.5)	…	(0,25,0,5,0,75)
C11	(0.5,0.75,1)	(0.5,0.75,1)	(0.5,0.75,1)	(0.5,0.75,1)	…	(0,25,0,5,0,75)
C12	(0.5,0.75,1)	(0.25,0.5,0.75)	(0.25,0.5,0.75)	(0,0.25,0.5)	…	(0,25,0,5,0,75)
C13	(0.25,0.5,0.75)	(0.5,0.75,1)	(0.5,0.75,1)	(0.7,1,1)	…	(0,5,0,75,1)
C14	(0.25,0.5,0.75)	(0.25,0.5,0.75)	(0.5,0.75,1)	(0.25,0.5,0.75)	…	(0,5,0,75,1)
C15	(0.5,0.75,1)	(0.25,0.5,0.75)	(0.5,0.75,1)	(0.25,0.5,0.75)	…	(0,0,25,0,5)
C16	(0.25,0.5,0.75)	(0.5,0.75,1)	(0.5,0.75,1)	(0.25,0.5,0.75)	…	(0,25,0,5,0,75)
C17	(0,0.25,0.5)	(0.5,0.75,1)	(0.25,0.5,0.75)	(0.5,0.75,1)	…	(0,5,0,75,1)
C18	(0.25,0.5,0.75)	(0.25,0.5,0.75)	(0.5,0.75,1)	(0,0.25,0.5)	…	(0,0,0,25)
Min	(0,0,0.25)	(0,0,0.25)	(0,0,0.25)	(0,0,0.25)	…	(0,0,0,25)
Max	(0.5,0.75,1)	(0.5,0.75,1)	(0.7,1,1)	(0.7,1,1)	…	(0,5,0,75,1)

The matrix formed by the normalisation of the values in Table 12, utilising Equations (10)–(13), is presented in Table 13.

Table 13. Normalised triangular fuzzy values for E1.

Criteria	C1	C2	C3	C4	…	C18
C1	(0,0,0)	(0,5,0,75,0,75)	(0.25,0.5,0.5)	(0.5,0.75,0.75)	…	(0,25,0,5,0,5)
C2	(0,5,0,75,0,75)	(0,0,0)	(0.5,0.75,0.75)	(0.25,0.5,0.5)	…	(0,25,0,5,0,5)
C3	(0,25,0,5,0,5)	(0,5,0,75,0,75)	(0,0,0)	(0.25,0.5,0.5)	…	(0,5,0,75,0,75)
C4	(0,25,0,5,0,5)	(0,5,0,75,0,75)	(0.5,0.75,0.75)	(0,0,0)	…	(0,5,0,75,0,75)
C5	(0,25,0,5,0,5)	(0,25,0,5,0,5)	(0.5,0.75,0.75)	(0.25,0.5,0.5)	…	(0,25,0,5,0,5)
C6	(0,0,25,0,25)	(0,25,0,5,0,5)	(0.25,0.5,0.5)	(0,0.25,0.25)	…	(0,0,25,0,25)
C7	(0,25,0,5,0,5)	(0,25,0,5,0,5)	(0,0.25,0.25)	(0.25,0.5,0.5)	…	(0,25,0,5,0,5)
C8	(0,25,0,5,0,5)	(0,0,25,0,25)	(0.25,0.5,0.5)	(0.25,0.5,0.5)	…	(0,0,25,0,25)
C9	(0,25,0,5,0,5)	(0,25,0,5,0,5)	(0.25,0.5,0.5)	(0,0.25,0.25)	…	(0,5,0,75,0,75)
C10	(0,25,0,5,0,5)	(0,5,0,75,0,75)	(0.7,1,0.75)	(0,0.25,0.25)	…	(0,25,0,5,0,5)
C11	(0,5,0,75,0,75)	(0,5,0,75,0,75)	(0.5,0.75,0.75)	(0.5,0.75,0.75)	…	(0,25,0,5,0,5)
C12	(0,5,0,75,0,75)	(0,25,0,5,0,5)	(0.25,0.5,0.5)	(0,0.25,0.25)	…	(0,25,0,5,0,5)
C13	(0,25,0,5,0,5)	(0,5,0,75,0,75)	(0.5,0.75,0.75)	(0.7,1,0.75)	…	(0,5,0,75,0,75)
C14	(0,25,0,5,0,5)	(0,25,0,5,0,5)	(0.5,0.75,0.75)	(0.25,0.5,0.5)	…	(0,5,0,75,0,75)
C15	(0,5,0,75,0,75)	(0,25,0,5,0,5)	(0.5,0.75,0.75)	(0.25,0.5,0.5)	…	(0,0,25,0,25)
C16	(0,25,0,5,0,5)	(0,5,0,75,0,75)	(0.5,0.75,0.75)	(0.25,0.5,0.5)	…	(0,25,0,5,0,5)
C17	(0,0,25,0,25)	(0,5,0,75,0,75)	(0.25,0.5,0.5)	(0.5,0.75,0.75)	…	(0,5,0,75,0,75)
C18	(0,25,0,5,0,5)	(0,25,0,5,0,5)	(0.5,0.75,0.75)	(0,0.25,0.25)	…	(0,0,0)

The right and left normalised values for the criteria are obtained by applying Equations (14) and (15) to the values specified in Table 13. These values are shown in Table 14. Crisp and total normalised crisp values were found using Equations (16) and (17), respectively. Total normalised crisp is shown in Table 15.

Table 14. Left and right normalised values for E1.

Criteria	C1	C2	C3	C4	…	C18
C1	(0,0)	(0.6,0.75)	(0.4,0.5)	(0.6,0.75)	…	(0.4,0.5)
C2	(0.6,0.75)	(0,0)	(0.6,0.75)	(0.4,0.5)	…	(0.4,0.5)
C3	(0.4,0.5)	(0.6,0.75)	(0,0)	(0.4,0.5)	…	(0.6,0.75)
C4	(0.4,0.5)	(0.6,0.75)	(0.6,0.75)	(0,0)	…	(0.6,0.75)
C5	(0.4,0.5)	(0.4,0.5)	(0.6,0.75)	(0.4,0.5)	…	(0.4,0.5)
C6	(0.2,0.25)	(0.4,0.5)	(0.4,0.5)	(0.2,0.25)	…	(0.2,0.25)
C7	(0.4,0.5)	(0.4,0.5)	(0.2,0.25)	(0.4,0.5)	…	(0.4,0.5)
C8	(0.4,0.5)	(0.2,0.25)	(0.4,0.5)	(0.4,0.5)	…	(0.2,0.25)
C9	(0.4,0.5)	(0.4,0.5)	(0.4,0.5)	(0.2,0.25)	…	(0.6,0.75)
C10	(0.4,0.5)	(0.6,0.75)	(0.77,1)	(0.2,0.25)	…	(0.4,0.5)
C11	(0.6,0.75)	(0.6,0.75)	(0.6,0.75)	(0.6,0.75)	…	(0.4,0.5)
C12	(0.6,0.75)	(0.4,0.5)	(0.4,0.5)	(0.2,0.25)	…	(0.4,0.5)
C13	(0.4,0.5)	(0.6,0.75)	(0.6,0.75)	(0.77,1)	…	(0.6,0.75)
C14	(0.4,0.5)	(0.4,0.5)	(0.6,0.75)	(0.4,0.5)	…	(0.6,0.75)
C15	(0.6,0.75)	(0.4,0.5)	(0.6,0.75)	(0.4,0.5)	…	(0.2,0.25)
C16	(0.4,0.5)	(0.6,0.75)	(0.6,0.75)	(0.4,0.5)	…	(0.4,0.5)
C17	(0.2,0.25)	(0.6,0.75)	(0.4,0.5)	(0.6,0.75)	…	(0.6,0.75)
C18	(0.4,0.5)	(0.4,0.5)	(0.6,0.75)	(0.2,0.25)	…	(0,0)

Table 15. Total normalised crisp values for E1.

Criteria	C1	C2	C3	C4	C5	C6	C7	C8	…	C18
C1	-	0.70	0.45	0.70	0.45	0.70	0.96	0.70	…	0.45
C2	0.70	-	0.70	0.45	0.70	0.45	0.70	0.70	…	0.45
C3	0.45	0.70	-	0.45	0.45	0.70	0.70	0.70	…	0.70
C4	0.45	0.70	0.70	-	0.45	0.45	0.96	0.70	…	0.70
C5	0.45	0.45	0.70	0.45	-	0.70	0.70	0.70	…	0.45
C6	0.21	0.45	0.45	0.21	0.21	-	0.45	0.45	…	0.21
C7	0.45	0.45	0.21	0.45	0.21	0.45	-	0.45	…	0.45
C8	0.45	0.21	0.45	0.45	0.45	0.21	0.70	-	…	0.21
C9	0.45	0.45	0.45	0.21	0.21	0.21	0.70	0.45	…	0.70
C10	0.45	0.70	0.96	0.21	0.21	0.70	0.96	0.45	…	0.45
C11	0.70	0.70	0.70	0.70	0.70	0.45	0.96	0.96	…	0.45
C12	0.70	0.45	0.45	0.21	0.21	0.45	0.70	0.70	…	0.45
C13	0.45	0.70	0.70	0.96	0.70	0.70	0.70	0.70	…	0.70
C14	0.45	0.45	0.70	0.45	0.70	0.70	0.96	0.96	…	0.70
C15	0.70	0.45	0.70	0.45	0.70	0.70	0.96	0.70	…	0.21
C16	0.45	0.70	0.70	0.45	0.70	0.45	0.96	0.96	…	0.45
C17	0.21	0.70	0.45	0.70	0.45	0.70	0.70	0.70	…	0.70
C18	0.45	0.45	0.70	0.21	0.70	0.45	0.45	0.45	…	-

The total normalised crisp values calculated for all experts are gathered and aggregated using Equation (18). These operations create the direct relation matrix shown in Table 16.

Table 16. Direct relation matrix.

Criteria	C1	C2	C3	C4	C5	C6	C7	C8	…	C18
C1	-	0.70	0.70	0.64	0.70	0.48	0.89	0.60	…	0.49
C2	0.61	-	0.79	0.70	0.76	0.67	0.73	0.70	…	0.55
C3	0.51	0.83	-	0.61	0.64	0.80	0.73	0.86	…	0.34
C4	0.51	0.67	0.67	-	0.57	0.58	0.67	0.60	…	0.37
C5	0.48	0.57	0.64	0.48	-	0.71	0.80	0.64	…	0.48
C6	0.27	0.54	0.54	0.48	0.43	-	0.73	0.61	…	0.36
C7	0.43	0.54	0.45	0.45	0.48	0.58	-	0.64	…	0.51
C8	0.54	0.55	0.52	0.48	0.55	0.46	0.64	-	…	0.25
C9	0.54	0.55	0.64	0.46	0.39	0.55	0.70	0.51	…	0.48
C10	0.67	0.70	0.83	0.61	0.67	0.80	0.73	0.73	…	0.57
C11	0.52	0.48	0.57	0.61	0.61	0.70	0.80	0.80	…	0.61
C12	0.42	0.70	0.70	0.61	0.64	0.77	0.67	0.60	…	0.42
C13	0.55	0.67	0.64	0.70	0.65	0.62	0.76	0.70	…	0.46
C14	0.67	0.67	0.80	0.61	0.64	0.80	0.89	0.86	…	0.39
C15	0.64	0.49	0.73	0.67	0.76	0.55	0.76	0.73	…	0.51
C16	0.51	0.64	0.53	0.60	0.54	0.67	0.80	0.76	…	0.51
C17	0.25	0.43	0.28	0.50	0.48	0.57	0.64	0.58	…	0.57
C18	0.64	0.67	0.61	0.28	0.73	0.67	0.67	0.76	…	-

The values of the direct relationship matrix are then normalised by dividing them by the largest value of the row and column totals of the same matrix (see Table 17). In the subsequent step, an 18 × 18 unit matrix is created and the normalised direct matrix values are subtracted from this matrix and inverted. The inverse matrix is demonstrated in Table 18.

Table 17. Normalised direct relation matrix.

Criteria	C1	C2	C3	C4	C5	C6	C7	C8	…	C18
C1	-	0.06	0.06	0.05	0.06	0.04	0.07	0.05	…	0.04
C2	0.05	-	0.06	0.06	0.06	0.05	0.06	0.06	…	0.04
C3	0.04	0.07	-	0.05	0.05	0.06	0.06	0.07	…	0.03
C4	0.04	0.05	0.05	-	0.05	0.05	0.05	0.05	…	0.03
C5	0.04	0.05	0.05	0.04	-	0.06	0.06	0.05	…	0.04
C6	0.02	0.04	0.04	0.04	0.03	-	0.06	0.05	…	0.03
C7	0.03	0.04	0.04	0.04	0.04	0.05	-	0.05	…	0.04
C8	0.04	0.04	0.04	0.04	0.04	0.04	0.05	-	…	0.02
C9	0.04	0.04	0.05	0.04	0.03	0.04	0.06	0.04	…	0.04
C10	0.05	0.06	0.07	0.05	0.05	0.06	0.06	0.06	…	0.05
C11	0.04	0.04	0.05	0.05	0.05	0.06	0.06	0.06	…	0.05
C12	0.03	0.06	0.06	0.05	0.05	0.06	0.05	0.05	…	0.03
C13	0.04	0.05	0.05	0.06	0.05	0.05	0.06	0.06	…	0.04
C14	0.05	0.05	0.06	0.05	0.05	0.06	0.07	0.07	…	0.03
C15	0.05	0.04	0.06	0.05	0.06	0.04	0.06	0.06	…	0.04
C16	0.04	0.05	0.04	0.05	0.04	0.05	0.06	0.06	…	0.04
C17	0.02	0.03	0.02	0.04	0.04	0.05	0.05	0.05	…	0.05
C18	0.05	0.05	0.05	0.02	0.06	0.05	0.05	0.06	…	-

Table 18. Inversing matrix.

Criteria	C1	C2	C3	C4	C5	C6	C7	C8	…	C18
C1	1.13	0.21	0.21	0.19	0.21	0.20	0.25	0.22	…	0.16
C2	0.19	1.17	0.24	0.21	0.23	0.23	0.26	0.25	…	0.17
C3	0.18	0.22	1.16	0.19	0.21	0.23	0.25	0.24	…	0.15
C4	0.16	0.19	0.19	1.13	0.18	0.19	0.22	0.20	…	0.14
C5	0.16	0.19	0.19	0.17	1.14	0.20	0.23	0.21	…	0.15
C6	0.12	0.16	0.16	0.15	0.15	1.12	0.20	0.18	…	0.12
C7	0.14	0.17	0.16	0.15	0.16	0.18	1.15	0.19	…	0.14
C8	0.14	0.15	0.15	0.14	0.15	0.15	0.18	1.12	…	0.11
C9	0.15	0.17	0.18	0.15	0.16	0.18	0.21	0.18	…	0.14
C10	0.19	0.22	0.23	0.19	0.21	0.23	0.25	0.24	…	0.17
C11	0.17	0.20	0.21	0.19	0.20	0.22	0.25	0.24	…	0.17
C12	0.15	0.19	0.19	0.17	0.18	0.20	0.22	0.20	…	0.14
C13	0.17	0.20	0.20	0.19	0.20	0.20	0.24	0.22	…	0.15
C14	0.18	0.21	0.22	0.19	0.20	0.22	0.26	0.24	…	0.15
C15	0.18	0.20	0.22	0.20	0.21	0.21	0.25	0.23	…	0.16
C16	0.16	0.19	0.19	0.18	0.18	0.20	0.23	0.22	…	0.15
C17	0.13	0.17	0.16	0.16	0.17	0.18	0.21	0.19	…	0.15
C18	0.18	0.20	0.20	0.16	0.21	0.21	0.23	0.23	…	1.12

The multiplication of this obtained matrix by the normalised direct matrix (see Equation (21)) gives rise to the total relation matrix, as presented in Table 19.

Table 19. Total relation matrix.

Criteria	C1	C2	C3	C4	C5	C6	C7	C8	C9	C10	…	C18
C1	0.13	0.21	0.21	0.19	0.21	0.20	0.25	0.22	0.19	0.21	…	0.16
C2	0.19	0.17	0.24	0.21	0.23	0.23	0.26	0.25	0.20	0.24	…	0.17
C3	0.18	0.22	0.17	0.19	0.21	0.23	0.25	0.25	0.18	0.23	…	0.15
C4	0.16	0.19	0.19	0.13	0.18	0.19	0.22	0.20	0.15	0.20	…	0.14
C5	0.16	0.19	0.19	0.17	0.14	0.20	0.23	0.21	0.16	0.19	…	0.15
C6	0.12	0.16	0.16	0.15	0.15	0.12	0.20	0.18	0.13	0.16	…	0.12
C7	0.14	0.17	0.17	0.15	0.16	0.18	0.15	0.19	0.15	0.18	…	0.14
C8	0.14	0.15	0.15	0.14	0.15	0.15	0.18	0.12	0.13	0.15	…	0.10
C9	0.15	0.17	0.18	0.15	0.16	0.18	0.21	0.18	0.11	0.18	…	0.14
C10	0.19	0.22	0.23	0.20	0.21	0.23	0.25	0.24	0.18	0.17	…	0.17
C11	0.17	0.20	0.21	0.19	0.20	0.22	0.25	0.24	0.18	0.21	…	0.17
C12	0.15	0.19	0.20	0.17	0.18	0.20	0.22	0.20	0.15	0.20	…	0.14
C13	0.17	0.20	0.20	0.19	0.20	0.21	0.24	0.22	0.18	0.21	…	0.15
C14	0.18	0.21	0.22	0.19	0.20	0.23	0.26	0.24	0.19	0.22	…	0.15
C15	0.18	0.20	0.22	0.20	0.21	0.21	0.25	0.23	0.18	0.23	…	0.16
C16	0.16	0.20	0.19	0.18	0.19	0.20	0.24	0.22	0.16	0.20	…	0.15
C17	0.13	0.17	0.16	0.16	0.17	0.19	0.21	0.19	0.15	0.19	…	0.15
C18	0.18	0.20	0.20	0.16	0.21	0.21	0.23	0.23	0.18	0.22	…	0.12

The row totals of the matrix shown in Table 19 provide the $R_i$ values, while the column totals provide the $C_j$ values. In addition, $(R_i+C_j)$ is the importance and $(R_i-C_j)$ is the net effect of criteria $i$.

The obtained values and identities of the criteria are shown in Table 20. The calculated threshold value, which is the average of the total relation matrix values, is 0.1757.

Table 20. Calculations of effect and cause values.

Criteria	${\mathit{R}}_{\mathit{i}}$	${\mathit{C}}_{\mathit{j}}$	${\mathit{R}}_{\mathit{i}}+{\mathit{C}}_{\mathit{j}}$	${\mathit{R}}_{\mathit{i}}$$- {\mathit{C}}_{\mathit{j}}$	Identity
Customisation (C1)	3.38	2.89	6.27	0.48	Cause
Stability (C2)	3.76	3.43	7.18	0.33	Cause
Functionality (C3)	3.51	3.49	7.00	0.02	Cause
Provider Support (C4)	3.03	3.13	6.16	−0.10	Result
Root Cause Analysis (C5)	3.08	3.36	6.44	−0.27	Result
Environmental Sustainability Tracking (C6)	2.56	3.60	6.15	−1.04	Result
Total Cost of Ownership (C7)	2.77	4.10	6.87	−1.33	Result
User-Friendliness (C8)	2.42	3.82	6.24	−1.40	Result
Maintainability (C9)	2.79	2.94	5.74	−0.15	Result
External Integration Capability (C10)	3.57	3.61	7.17	−0.04	Result
AI-powered optimisation and decision support (C11)	3.43	3.04	6.47	0.40	Cause
Deployment Type (C12)	3.01	2.98	5.99	0.03	Cause
Scalability (C13)	3.31	2.99	6.30	0.32	Cause
Simulation Capability (C14)	3.43	2.83	6.26	0.60	Cause
Internal Integration and automation capability (C15)	3.46	2.76	6.22	0.71	Cause
Real time monitoring (C16)	3.18	2.77	5.94	0.41	Cause
Cyber Security Posture (C17)	2.91	2.58	5.50	0.33	Cause
Compliance With Regulatory Frameworks (C18)	3.32	2.60	5.92	0.72	Cause

As demonstrated in Table 20, the influenced factors are C4, C5, C6, C7, C8, C9 and C10, while the others are influential factors. Furthermore, the top three criteria with the highest overall importance in the system are C2, C10 and C3, respectively. In this context, while the various features of the TOS are important, their effectiveness is contingent on the stability of the system. Moreover, the fact that integration with external stakeholders ranks second can be linked to the importance of container terminals’ integration into supply chains. The significance of functionality suggests that TOS is anticipated to be employed across a range of domains and processes within the terminal. C2 is the most influential factor, while C7 is the most influenced criterion. It is important to note that costs are inherently volatile, sensitive and dependent on a number of other factors. Within this paradigm, the magnitude of these effects is directly proportional to the augmentation of the system’s capabilities.

The causal relations map of criteria, created according to the values specified in Table 20, is shown in Figure 4.

Figure 4. Causal relations map of criteria.

The threshold value and strength of the interactions between criteria are of great importance for interpreting the obtained values. Figure 5 was created using the total relation matrix, with the threshold value as a reference. The strength of the interactions was categorised through division of the range from the threshold (0.1757) to the maximum value (0.2627) of total relation matrix. According to this, the following applies:

$$\mathrm{If} = \left\{\begin{array}{c}{t}_{ij}\le 0.1757⇨ \mathrm{no} \mathrm{significant} \mathrm{interaction} (\mathrm{visualised} \mathrm{no} \mathrm{color})\\ {0.1757<t}_{ij}\le 0.2044 ⇨ \mathrm{weak} \mathrm{interaction} (\mathrm{visualised} \mathrm{light} \mathrm{blue})\\ {0.2044<t}_{ij}\le 0.2335 ⇨ \mathrm{moderate} \mathrm{interaction} (\mathrm{visualised} \mathrm{medium} \mathrm{blue})\\ {0.2335<t}_{ij}\le 0.2627 ⇨ \mathrm{strong} \mathrm{interaction} (\mathrm{visualised} \mathrm{dark} \mathrm{blue})\end{array}\right.$$

Figure 5. Strength of the interactions between criteria.

Consequently, the C1, C2, C3, C10, C11, C13, C14, C15 and C18 criteria have a strong influence on C7. In a similar manner, C2, C3, C10, C11, C14 and C15 have been demonstrated to exert significant influence on C8. It is evident that the system’s fundamental functions and its compliance with regulatory frameworks, in conjunction with the optimisation, integration and advanced technology capabilities of the TOS, have a considerable impact on costs and user-friendliness. Furthermore, it is evident that stability exerts a substantial influence on functionality and external integration capability. In conclusion, the findings indicate that the system’s features and capabilities will remain theoretical as long as they remain unrealised and interrupted. In this context, although TOS features such as stability, functionality, integration, optimisation and simulation will particularly influence terminal operators’ investment decisions and increase costs, they will enhance the advantages provided to users.

The weights and causal centrality-based graphical representation, created in accordance with the findings obtained using the FAHP and F-DEMATEL methods, is shown in Figure 6. The x-axis of the graph is composed of ($R_i$ − $C_j$) values, while the y-axis is composed of ($w_i$) values. The reference point for the y-axis is 0.055, which is the average of the criterion weights. The delineation of the regions constituting the graph, along with the pertinent criteria, is elucidated in Table 21.

Figure 6. Quadrant graph based on weights and causal centrality of criteria.

Table 21. Definitions of quadrant regions.

Quadrants	Explanations	Criteria
1	High priority and positive causality	C2, C3, C11, C14, C15, C16, C18
2	High priority and negative causality	C7, C10
3	Low priority and negative causality	C4, C5, C6, C8, C9
4	Low priority and positive causality	C1, C12, C13, C17

The selection of a TOS is a management decision that is linked to operational strategy rather than an IT decision. Incorrect decisions in this issue, characterised by a substantial financial investment, have the potential to precipitate a sequence of challenging complications. As demonstrated in Table 21 and Figure 6, from the perspective of terminal operators, TOS investment decisions are predominantly influenced by stability, functionality, AI-based decision support, simulation, internal integration, real-time monitoring and legal compliance. These criteria are fundamental to the success of TOS investment decisions. TCO and external integration are more influenced by these criteria and improve as the cause criteria are enhanced. In the third quarter, these criteria are of lesser importance than other criteria and are subject to influence. While these factors constitute a secondary group of criteria in investment decisions, they ought to be evaluated as performance outputs linked to the cause criteria. The criteria in the fourth quarter are of relatively low importance, have a causal effect, and are the triggering characteristics of the system’s success.

6. Discussion and Conclusions

The volume of international trade, increasing cargo volumes, growing ship sizes and complex logistics networks have made the management of port processes even more challenging. In this context, the utilisation of advanced technology-based systems has become imperative for seaports, which are pivotal to supply chains. Despite the majority of ports yet to achieve a high level of digital maturity, Industry 4.0 technologies are changing port business models. It is particularly noteworthy that a number of prominent ports, both large and medium-sized, are allocating substantial resources to the development of infrastructure, technological equipment, information technologies and automation systems. These investments are being made with a view to achieving environmental sustainability, digitalisation, integration, modernisation, safety and security. In this context, Terminal Operating Systems (TOSs) play a pivotal role in the management of all operations and systems utilised. Comprising numerous modules and usable for different types of cargo terminals, TOSs are defined as the brain of ports and are considered to be strategic for advanced port operations. In view of the information presented, a comprehensive set of TOS selection criteria (18 criteria) for container terminals has been established and their importance levels have been calculated. Furthermore, the interactions between the criteria have been determined in the present study. To the author’s knowledge, this is the first study conducted on this subject with this scope. The establishment of the set of criteria was informed by academic literature and expert opinions. In this particular context, the study was conducted by a team of eight experts directly related to the subject matter, which significantly strengthens the reliability and applicability of the methodology.

The defined TOS selection criteria set was weighted using FAHP and the interactions between these criteria were examined using F-DEMATEL. The findings indicate that, in addition to certain fundamental characteristics (stability, functionality, TCO and compliance with legal requirements) expected from all information systems, internal and external integration, automation, real-time monitoring, AI-powered optimisation and simulation capabilities are of similar importance. Indeed, factors such as these take precedence over user-friendliness, provider support and customisation. This suggests that TOSs should be structured in a manner that facilitates the smartisation process of ports and enhances their digital maturity. In this particular context, it is evident that container terminal operators anticipate a TOS structure that can collaborate effectively with advanced technologies and also provide automation and decision support. Furthermore, the fact that TOSs should support real-time monitoring and integration with external stakeholders is also critical in terms of collaboration and transparency, which are of great importance in supply chain management. The findings of this study are consistent with and parallel to those of the studies by [5,18,57,62] A further crucial point pertains to the rankings of root cause analysis and environmental sustainability tracking criteria. The rankings indicate that terminal operators anticipate TOS support for the issues they face in identifying the root causes of specific problems, resolving them and monitoring environmental sustainability. It is interesting to note that cybersecurity posture is one of the relatively less important criteria. Terminals are not merely facilities for cargo transfers; they are also sites for intensive data transfer. In addition, it is important to note that terminal systems have the capacity to function in conjunction with a multitude of systems for the purpose of data sharing. This renders terminals susceptible to cyber threats. The relatively low importance value of this criterion indicates that, although experts consider TOS cyber security important, they view basic protection against cyber threats as the terminal’s primary multi-layered security architecture. It is evident that the cost criterion is a pivotal factor in investment decision-making. The findings of the present study lend support to this hypothesis. This finding is also consistent with the studies by [4,18]. Furthermore, it has been determined that the cost criterion is the most affected by other criteria. The findings of the F-DEMATEL analysis indicate that the integration with external stakeholders is of paramount importance, once again emphasising the pivotal role these terminals play in supply chains. Terminal operators anticipate that TOSs should contribute to the mitigation of specific deficiencies within this domain. In conclusion, container terminal operators expect TOSs to integrate complex business processes, provide effective decision support, enhance traceability, work seamlessly with advanced technologies and external systems and enable rapid intervention in bottlenecks.

The present study is expected to contribute to the TOS literature and provide benefits to container terminal managers in making more effective TOS investment decisions, thereby gaining a competitive advantage. It is hypothesised that the study will furnish port authorities and public institutions in the sector with the necessary insight to facilitate the development of effective port digitalisation policies. Furthermore, it is hoped that it will provide a framework for TOS providers to better understand their customers’ expectations and prioritise their product designs by examining the interactions of different criteria. Nevertheless, the present study is subject to various limitations. The analyses in the relevant study were conducted based on the assessments of experts located in Turkey who were easily accessible and the study is subject to geographical limitations. In this context, the study exclusively encompasses the opinions of experts in Turkey, thereby restricting its generalisability. It is important to note that the results obtained from interviews and surveys conducted with experts located in different countries may vary. In addition, it is anticipated that TOSs to be utilised in autonomous ports, which are projected to be managed exclusively by software and machinery in the near future, will be developed. These TOSs, which are likely to be based on deep learning, will offer more advanced functions and respond to different needs. In this context, the criteria to be determined will change and expand. In future studies, it is recommended that evaluations be conducted utilising novel sets of criteria and methodologies. In addition, the evaluation of TOSs currently in use in ports, using various sets of criteria or the performance of cost–benefit analyses, will also contribute to the extant literature. While the present study could also be of benefit to terminals handling other types of cargo, it has been conducted specifically for container terminals. In this context, further studies on other types of cargo (e.g., liquid bulk cargo, dry bulk cargo, neo-bulks) terminals would also be beneficial. It is evident that the efficiency and effectiveness of container and other cargo terminals, as a result of their investment in these systems and technologies, is of significant importance to ports, stakeholders and national economies. However, it should be noted that such investments require substantial financial resources. In this context, the provision of support through the implementation of incentives is likely to facilitate this process. It is imperative to acknowledge that, despite the technological underpinnings of these advancements, the demand for qualified and skilled workers remains paramount.