Review and Classification of Objectives in Dynamic Dial-a-Ride Systems: A Triple Bottom Line Approach of Sustainability

Abstract

Dynamic dial-a-ride problems (DDARPs) involve designing routes and schedules for customers with specific origins and destinations. While the optimization of DDARPs has been extensively examined, these analyses often focus solely on economic decisions. The recent literature emphasizes the inclusion of social and environmental factors in addition to economic considerations for a sustainable transportation system. This paper provides a conceptual review that identifies and classifies the most common DDARP objectives in the three dimensions of the Triple-Bottom-Line (3BL) approach of sustainability: environmental, economic, and social. This study analyzes the interconnections among different objectives and provides insights into multi-objective approaches used in transportation problems. The findings demonstrate the interconnectedness of objectives from different dimensions and highlight the involvement of various stakeholders in decision-making. The results show that optimizing one objective may have implications for other objectives, suggesting a trade-off to be considered. The results reveal that social objectives boost the economic dimension by improving service quality; however, environmental objectives negatively impact the economic dimension. Additionally, a geographical analysis was conducted, which revealed continent-wise variations in research focus and contributions. Future studies should focus more on the social and environmental dimensions to promote a sustainable transportation system.

1. Introduction

In recent years, the surge in popularity of ride-hailing services has attracted increased attention to the vehicle routing problem (VRP) and its diverse variants. The VRP involves optimizing vehicle routes and schedules to accommodate customers with specific origins, destinations, and departure times. Different variants of the VRP address realistic scenarios with distinct constraints and objectives, such as the VRP with time windows (VRPTW) and the pickup and delivery problem with time windows (PDPTW). However, the dial-a-ride problem (DARP) distinguishes itself from similar routing problems by emphasizing the human factor. In transporting human passengers, minimizing user inconvenience becomes crucial, especially considering the limited vehicle capacity compared to parcel or letter delivery problems. Ho et al. [1] proposed a taxonomy for classifying DARPs based on two key dimensions: evolution and quality of information. The evolution of information relates to changes in available data throughout the execution of routes, while the quality of information relates to the potential uncertainty associated with the data, such as when customer demand is estimated within a range instead of an exact value. The dynamic nature of the problem requires the real-time design of vehicle routes and communication with the next customer to be served, referred to as the dynamic DARP (DDARP). This study focuses on the research developments in DDARPs since 2007, building upon the previous review by Cordeau and Laporte [2].

Section 1.1 introduces the main features of DDARPs, covering users, the system, and related terminologies. Section 1.2 explores the diverse application areas of DDARPs, including services for the elderly and disabled, which is the conventional use of a dial-a-ride (DAR) service. Section 1.3 outlines the motivation for this study, emphasizing the importance of surveying to enhance understanding of sustainable approaches in solving DDARPs.

1.1. Features of DARPs

In DDARP, multiple users request transportation between two specified locations (pickup and drop-off). The service provider receives the requests and schedules their fleet to serve the requests. In the case of ride-sharing transportation services, there can be more than one user in the vehicle simultaneously, which can result in detours for some users. The typical features of a DDARP are discussed below:

The DDARP is characterized by its dynamic nature, with the problem constantly evolving as new requests are received and scheduled. Users are the entities that need transportation, and they can be either people or goods, while the fleet is the number of vehicles available for transportation service. Vehicles complete a trip once they return to the depot, and a single vehicle may complete multiple trips in a day. Vehicle capacity is essential for passenger transportation services, as it determines the maximum number of users that can be transported simultaneously. Ride time is the time users spend in a vehicle, which is the difference between the pickup and drop-off times.

The DDARP is solved by optimizing an objective function from economic to non-economic considerations. The problem has several constraints, the most important being the time window constraints. These constraints specify the earliest and latest times for pickup and drop-off as specified by the user and can be either soft or hard constraints based on user requirements.

1.2. Application Areas of DARPs

DAR services traditionally catered to the elderly and disabled population, aiming to fulfill demand while minimizing operational costs [3,4]. DAR services are also extensively utilized in the health sector, specifically for addressing emergencies and ensuring compatibility between staff and equipment. In this context, prioritizing requests based on urgency and constraints is crucial to maximizing patient convenience [5,6]. The integration of demand-responsive transport (DRT) with public transportation (PT) is an emerging area in DAR services, providing an alternative during low-demand periods, specific locations, or when PT services are unavailable [7,8]. Private companies operate these services with the primary objective of minimizing costs and maximizing profits [9,10]. Dynamic ride-sharing problems arise when DRT enables ride-sharing among users, resulting in increased dynamism and complexity of the problem [11,12,13,14]. More details on the application areas of DARPs can be found in Ho et al. [1].

1.3. Motivation for the Review

The increasing popularity of DAR services is accompanied by concerns regarding their impact on the environment and society. Organizing a DAR service can offer significant economic benefits to providers and users while also affecting social welfare and the environment [15]. While these services offer accessible and efficient mobility options, achieving sustainable development remains challenging due to conflicting objectives of cost-effectiveness, high-quality service, and environmental/social impact. Black [16] defines sustainable transportation as fulfilling current needs without impeding future generations’ capacity to meet their needs. This definition aligns transportation with the “Triple Bottom Line” (3BL) objectives for sustainability [17], which are the combination of the economic, environmental, and social dimensions [18].

Various sustainability assessment methods are studied in the literature, focusing on one or more dimensions. For instance, Life-Cycle Assessment (LCA) primarily targets environmental impacts but often overlooks social and economic dimensions [19]. Cost–Benefit Analysis (CBA) emphasizes economic efficiency but fails to capture social and environmental benefits and costs [20]. Social Return on Investment (SROI) focuses on social impacts by translating social outcomes into monetary values but may not fully integrate environmental and economic dimensions [21]. Human and Social Capital Measurement (HSCM) assesses the social benefits of a project by measuring improvements in human and social capital but does not consider economic and environmental benefits [22]. Most of these approaches ignore the simultaneous consideration of all three dimensions, economic, social, and environmental, creating a gap in comprehensive sustainability assessments. This study aims to bridge this gap by adopting the 3BL approach in the field of transportation, integrating all three pillars of sustainability. By classifying the objectives of DDARPs using the 3BL approach, this review seeks to align them with different sustainability dimensions and contribute to achieving sustainable development in DAR services.

The rest of this paper is organized as follows: Section 2 describes the methodology for paper screening and the characteristics of the selected papers. Section 3 explores the 3BL approach and its correlation with the objectives. Section 4 explains the type of DDARPs and the general framework used in this study to classify different objectives. Section 5, Section 6 and Section 7 review the economic, social, and environmental objectives, respectively. Section 8 evaluates the study and provides a geographical analysis followed by the implications of this study. Section 9 concludes this paper, discussing limitations and recommendations for future research.

2. Scope of the Survey and Methodology

This section outlines the methodology employed for conducting the literature search and review, explaining the search framework and the characteristics of the selected studies.

2.1. Search Framework

A specific search and screening mechanism, illustrated in Figure 1, was followed to identify the targeted research papers for review. The relevant studies were found following the proposed systematic search and screening mechanism, which was designed in four steps. In Step (1), the Scopus search engine is used to extract the studies in this review paper’s scope. Hence, on 20 March 2023, several simple key terms such as “Dynamic Dial-a-Ride Problem”, “Dynamic Dial a Ride Problem”, “Dynamic Ridesharing”, and “Dynamic Ride-Sharing” were applied to find the studies. The other details, such as search domain, research language, source type, and research publication year, are illustrated in Step (1) of Figure 1. The search was conducted within the domains of “Title, Abstract, and Keywords” to ensure comprehensive coverage, and it was limited to articles published in English to maintain consistency. This initial search resulted in 190 studies. Given the conceptual nature of this review, only key terms related to DDARPs were used to provide a representative sample for review and to conceptualize these objectives.

In Step (2), the abstracts of the 190 studies are checked, and only the in-scope studies (114 papers) are selected for full-text and reference screenings. In Step (3), the full text and references of 84 papers were screened. Papers were included if they specifically addressed dynamic routing problems, employed optimization techniques, and discussed practical applications in transportation management. Papers that did not focus on routing these services were excluded from the review. In Step (4), additional relevant papers were identified by screening the citations of the selected papers, ensuring that the most cited peer-reviewed articles for each dimension were included. This resulted in a final set of 98 papers for review.

2.2. Characteristics of Selected Papers

This article is structured so that the DDARPs between 2007 and 2023 have been included. Later, the relevant papers were identified from the cited papers in this searched literature. This resulted in a final of 98 papers, which were used for the conceptual review in this paper. Figure 2 presents the year-wise distribution of these studies, which shows that 63% of studies were conducted in the last eight years.

Figure 3 illustrates the distribution of papers by country, indicating that this study contains papers from 27 countries spanning six continents. Figure 3 shows that the maximum number of papers are published in the USA (19 papers), followed by France (11 papers) and China (8 papers).

Figure 4 shows the distribution of the number of papers across the different continents. The analysis shows that the maximum number of papers are published in Europe (47 papers), followed by North America (22 papers), Asia (15 papers), Africa (6 papers), South America, and Oceania (4 papers each), respectively. When a research study involves multiple authors from various countries, the geographic location attributed to the paper is determined based on the corresponding author’s affiliation. This approach ensures consistency and provides a centralized representation of the study’s origin.

3. Triple Bottom Line Approach (3BL)

Jeurissen [17] introduced the 3BL approach, which incorporates sustainable development principles and enables progress across multiple dimensions: economy, society, and the environment. The 3BL approach, known as the 3Ps, stands for profit, people, and planet, representing economic, social, and environmental impacts, respectively. These dimensions are interconnected and must be considered together to achieve sustainability, as depicted in Figure 5 [23]. The figure shows that sustainability can only be achieved through the coordinated interaction of all three dimensions, which partially overlap.

The economic dimension of 3BL focuses on creating value beyond financial performance, such as its profit, and accounting for the economic and operational impacts of the system. In the context of transportation, economic sustainability is achieved when the system efficiently moves people and goods while avoiding inequalities in accessing transportation services and resources. Achieving economic sustainability in transportation is crucial because it impacts the financial performance of transportation organizations and users’ overall economic and social well-being. It leads to cost savings, which can be reinvested into the system to improve service quality and expand coverage areas.

While the economic dimension of 3BL has received more attention, the social dimension is often overlooked. The social dimension of 3BL includes the impact of a system on the welfare of its stakeholders, such as users, and addresses concerns like better accessibility and equity. Social sustainability in transportation requires meeting the travel needs of all individuals safely and consistently, providing equitable and accessible transportation options regardless of socioeconomic status, age, or ability. For example, enhancing transportation accessibility in low-income areas promotes social equity by ensuring all individuals can access essential services and economic opportunities. Improved transportation options in these areas can lead to job creation and economic growth, demonstrating a clear link between social and economic sustainability.

The environmental dimension relates to the system’s attempts to minimize the environmental impact of different services. To achieve environmental sustainability in transportation, the system must prioritize objectives such as reducing resource consumption and mitigating adverse effects on ecological systems, including emissions, noise, and pollution. Key factors such as vehicle distance traveled (VDT), average route length, and the number of vehicles must be carefully managed to minimize their environmental externalities. A comprehensive approach is needed to achieve environmental sustainability in transportation, including promoting active transport, optimizing vehicle use, and reducing emissions to minimize the overall environmental impact and create a sustainable future. This benefits the environment and can lead to long-term cost savings and improved public health outcomes, linking environmental sustainability with economic and social benefits.

All these dimensions are interdependent. For instance, optimizing routes for economic efficiency often reduces operational costs and fuel consumption, lowers emissions, and demonstrates an overlap with environmental sustainability. Enhancing transportation accessibility in low-income areas promotes social equity and stimulates local economic development by improving employment opportunities, highlighting the interrelation between social and economic dimensions. Implementing electric vehicles in a transportation fleet reduces emissions and operational costs over time, benefiting the environment and the economy. Additionally, improved air quality resulting from lower emissions enhances public health, linking environmental and social dimensions. A detailed discussion of these interrelations is presented in Section 8.3.

In summary, the transportation system should balance equitable and environmentally responsible practices with economic viability and societal well-being to maximize profits sustainably. Therefore, it is essential to balance the economic objectives of transportation systems with the social and environmental objectives to achieve sustainability in the transportation system. Adopting the 3BL framework for this study, we categorize the objectives of DDARPs into three dimensions: economic, social, and environmental. The objectives focusing on the service provider’s needs are considered under the economic dimension, the problem focusing on the users’ objectives are considered under the social dimension, and the objectives related to emission and fuel consumption are characterized under the environmental dimension.

4. Different Objectives of DDARPs

This section discusses the types of problems in DDARPs, specifically single-objective and multi-objective problems. Single-objective problems focus on optimizing a specific objective, while multi-objective problems aim to simultaneously satisfy more than one objective. The section also presents a general framework for objective classifications and provides an overview of the distribution of objectives in the selected papers.

4.1. Types of Problems

This section discusses the types of problems studied in the literature, focusing on one or more than one objective, known as single- and multiple-objective problems.

4.1.1. Single-Objective Problems

Single-objective DDARPs are the most studied type of DARP in the literature, and they involve optimizing a single objective, typically from one of the dimensions. This approach allows a detailed analysis of the chosen objective, which can focus on user perceptions, such as minimizing total inconvenience to users [24]; service providers’ perspectives, such as minimizing the total operational costs [13]; or the perspective of the environment, such as minimizing the vehicle emissions [12]. A detailed review of these studies is provided in the following sections. Focusing on one objective facilitates precise measurements and assessment of the results, which is helpful for the stakeholders aiming to solve a real-life problem. Nevertheless, this approach overlooks the interdependencies of the different objectives and aspects of the transportation system, where optimizing one object affects the others. In summary, while single-objective DDARPs offer valuable insights into the individual objectives of a problem, their limitations underscore the importance of considering a more balanced approach for a comprehensive service design.

4.1.2. Multi-Objective Problems

While a considerable proportion of problems in this field have centered on a single objective, there are also those that consider multiple objectives. In these scenarios, the objective is to generate an optimal solution that simultaneously satisfies various objectives. Ho et al. [1] classified multi-objective problems into three categories: using a weighted sum of objectives, employing lexicographic objective functions, and adopting the Pareto frontier approach. The weighted sum of different objectives uses different weights for each objective, and the optimal solution for these objectives is determined. This type of multi-objective problem is appropriate when the weights of different objectives have been well-defined and evaluated. This approach is not applicable when the relative importance of each objective is unquantifiable. When dealing with multi-objective problems, it is crucial to consider the trade-offs between objectives. In some cases, optimizing one objective may come at the expense of sacrificing another. Therefore, it is essential to balance these objectives to achieve an overall satisfactory solution. For instance, in some cases, the decision-maker may be risk-averse, and therefore, the Pareto Frontier approach may be more suitable as it provides the decision-maker with a set of optimal solutions that are not dominated by any objectives and allows them to choose based on their preferences. In contrast, the weighted sum approach is more suitable for decision-makers who prioritize specific objectives and can assign weights to each objective based on their importance. The lexicographic approach, on the other hand, is more appropriate for problems where there is a clear hierarchy of objectives, and some objectives are more critical than others. Furthermore, the choice of the optimization approach also depends on the problem’s nature and the decision-maker’s preferences.

4.2. Framework of Objectives Classification

Figure 6 illustrates the main objectives of DDARPs found in the literature and characterizes them in their relevant dimension. A detailed review of each objective is carried out in the following sections. Some objectives fall under more than one dimension, and Figure 6 classifies them into their primary and most relevant dimension. For the sake of simplicity, the objectives’ names have been written as the shorter versions of the main objectives.

Figure 7 shows the objective-wise distribution of the papers used in this study. The figure shows that the maximum number of papers focus on optimizing total operational costs in their study (29%), followed by optimizing user inconvenience and service quality (13.5%) and maximizing the number of served requests (10.5%).

5. Economic Dimension Objectives

The economic dimension of 3BL pursues to create value beyond financial performance by considering a system’s economic and operational impacts on the system. This dimension is the most frequently studied in the literature, with most studies focusing on maximizing service providers’ profit. It is crucial to recognize that the objectives related to the social and environmental dimensions also impact the economic dimension, as they influence the system’s overall operational costs [18]. For instance, minimizing travel time and wait time may have externalities on the economic dimension by affecting the operational costs. Similarly, reducing emissions, VDT, and energy consumption contributes to the environmental dimension while ultimately reducing operational costs. Economic sustainability in transportation can be achieved when the system efficiently moves people and goods while promoting equity and operational efficiency. These are the most common objectives in this dimension found in the literature.

5.1. Minimize the Number of Vehicles

Minimizing the number of vehicles used in DDARPs is a common objective for service providers to reduce operational costs, traffic congestion, and carbon footprint. By reducing vehicle usage, service providers can lower acquisition, maintenance, and operations expenses. This objective also contributes to the environmental dimension by reducing VDT, fuel consumption, and emissions. Additionally, minimizing the number of vehicles helps service providers cope with resource constraints and enhances their capacity to serve more customers effectively.

Wong et al. [25] explored the operating efficiency of a dispatching system with a degree of dynamism by using an insertion-based heuristic to solve the problem of minimizing the total number of vehicles. Spieser et al. [26] developed a fluid-based optimization framework for a shared mobility system with time-varying demand with the objective of achieving the minimum number of vehicles required to avoid passenger queueing. Escuín et al. [27] presented a cooperative scheduling algorithm for solving dynamic PDPTW (DPDPTW) routing and scheduling aspects to minimize the total number of vehicles. Vonolfen and Affenzeller [28] presented and evaluated diverse general and specialized waiting for heuristics for the dynamic pickup and delivery problems (DPDP), with the primary objective being minimizing the number of vehicles. Guerram [29] proposed a multi-agent architecture for DPDP where agents are organized by groups and roles and solve the problem by optimizing a multi-objective function whose primary objective was to minimize the total number of vehicles.

5.2. Minimize Vehicle Distance Traveled

Minimizing VDT has significant economic implications, as it reduces transportation costs for service providers and users. Minimizing VDT can also mitigate the negative impact of congestion and associated environmental costs. In DDARPs, prioritizing the minimization of VDT ensures efficient and prompt service for passengers while reducing costs [30].

Issaoui et al. [31] presented a multi-objective mathematical model to resolve the DAR approach based on non-dominated sorting genetic algorithm II (NSGAII), with part of the objectives being the minimization of total VDT. Sayarshad and Chow [32] proposed a new DDARP featuring non-myopic pricing based on optimal tolling of queues to minimize the total VDT. Deleplanque and Quilliot [33] studied a form of DDARP by integrating a measure of insertion capacity to minimize the total VDT. Wang et al. [34] provided multiple methods for generating stable or nearly stable ride-share matching solutions, with the primary objective being the maximization of the saved traveled miles. Kucharski and Cats [35] propose an exact, replicable, and demand-driven algorithm for matching trips into shared rides with the objective of minimizing the VDT. Silwal et al. [36] proposed a dynamic and hybrid model using a particle swarm optimization-based evolutionary heuristic algorithm for ride-sharing systems to minimize the total VDT.

5.3. Maximize the Number of Served Requests

Maximizing the number of served requests in DDARPs is a crucial objective that benefits service providers and passengers and the broader accessibility of transportation services. Operators can increase revenue, reduce costs, and improve service availability by serving more requests. This objective is essential for enhancing transportation accessibility and supporting vulnerable populations who may face limitations in using conventional transportation methods.

Karabuk [37] developed a nested column generation method for solving the paratransit vehicle scheduling problem where the overall objective was to maximize the number of customers served. Pavone et al. [38] presented adaptive algorithms for motion coordination of a group of autonomous vehicles (AVs) to maximize the total number of requests while minimizing the total system time. Häme and Hakula [39] developed a maximum cluster algorithm based on dynamic programming to solve the DDARP and maximize the number of customers served. Stiglic et al. [40] conducted an extensive computational study to quantify the impact of different types of participants on the performance of a ride-sharing system by maximizing the number of matches. Masoud and Jayakrishnan [41] formulated a multi-hop P2P ride-sharing problem as a binary program with the objective of maximizing the number of served riders while minimizing the number of transfers. Later, in another study, Masoud and Jayakrishnan [42] proposed a solution to the general peer-to-peer multiple matching problems with transfers in a time-dependent network to maximize the number of served rides. Vallee et al. [43] developed a real-time DDARP service operated by the company ‘PADAM’ to maximize the number of accepted requests. Kumar and Khani [44] solved a ride-matching algorithm for integrating peer-to-peer ride-matching with schedule-based PT for first/last mile (FMLM) access to maximize the number of matches. Molenbruch et al. [45] introduced a routing algorithm and integrated scheduling procedure to enforce the synchronization with PT in an integrated mobility system, aiming to serve all user requests while serving a few partly covered by PT. De Ruijter et al. [46] examined the trade-off between computing time and solution quality under realistic urban on-demand ride-sharing service for a DDARP. The service quality was measured by maximizing the number of served requests.

5.4. Minimize the Total Operational Costs

Minimizing total operational costs and maximizing total profit are crucial objectives for DDARPs, benefiting both service providers and passengers. By reducing expenses associated with vehicle maintenance, fuel consumption, and staffing, service providers can operate more efficiently and generate higher profits. This can lead to service expansion, affordable pricing, and improved passenger accessibility. However, achieving this objective requires a balance with other objectives. Reducing costs excessively may affect service quality and passenger satisfaction, while using larger vehicles can increase environmental impacts. On the other hand, minimizing operational costs can enhance social equity and support environmental objectives by making transportation services more affordable and enabling investments in sustainable infrastructure and technologies. Optimizing operational costs and profit maximization play a significant role in achieving economic sustainability while considering social and environmental implications. An optimal balance between operational costs and profit maximization is integral to achieving economic sustainability while considering social and environmental factors.

The studies that focus on dynamic ride-sharing and aim to minimize the total operating costs are [4,13,47,48,49,50]. Some studies [4,51] solved the DDARP for hospitals where the main objective was to serve all patients while minimizing the operational costs of the service. Various studies have used this as their primary objective in different applications, in the case of logistics [5,52], courier services [6], and parcel delivery [53]. Mourdjis et al. [54] minimized the service costs based on the distance and the driver salaries in the supermarket’s supply field. Multiple studies [55,56,57,58] solved different DPDPTW to minimize operational and transportation costs. Daoud et al. [10] proposed a decentralized protocol for an on-demand transportation service to minimize the total operating costs. Other studies focused on minimizing the operational costs of DRT [9], taxis [59], and paratransit [3,60]. Zhu et al. [61] minimized charging costs, considering the energy consumption of electric vehicles. Other studies [62,63] focused on vehicle matching with minimizing service costs. Su et al. [64] introduced a DDARP in the shared logistics platform with the objective of maximizing the total profit. Dong et al. [65] solved a classic DDARP with the objective of minimizing the total costs while accommodating all requests. Later, Dong et al. [66] introduced a novel chance-constrained DARP (CC-DARP) to capture users’ preferences while maximizing the total profit.

5.5. Minimize Total Ride Time

From the service provider’s perspective, minimizing total ride time is a crucial objective for DDARPs because it reduces total operational costs by lowering drivers’ fuel consumption, vehicle maintenance expenses, and labor costs. It improves efficiency by optimizing resource allocation, leading to better resource utilization and increased productivity. Additionally, minimizing ride times expands service capacity, accommodates more ride requests, and potentially generates additional revenue. By minimizing total ride time, service providers can reduce the time passengers spend in a vehicle, which can help improve passenger satisfaction and overall service quality, which can correlate with the social dimension of the 3BL. This objective also aligns with the environmental dimension because shorter ride times can lead to higher vehicle speeds and less congestion, resulting in reduced fuel costs [67].

Jia et al. [68] discussed a periodic and event-driven rolling horizon procedure method to minimize the total traveling time for solving the DPDPTW. Hyytiä et al. [69] considered a dynamic VRP (DVRP) with pickup and delivery to minimize the total ride duration. Schilde et al. [70] compared four different metaheuristics to solve the DDARP to minimize the total tardiness and number of routes. Jung et al. [71] proposed three approaches to optimize the vehicle schedules of a DDARP, focusing on minimizing the passenger ride time and maximizing the provider’s profit. Yang et al. [72] presented a multi-objective memetic algorithm based on request prediction for route planning where two main objectives, route length and travel time, are optimized. Levin et al. [73] proposed an event-based framework for implementing shared AVs’ behavior in existing traffic simulation models to minimize the user’s travel time.

6. Social Dimension Objectives

The social dimension, the second dimension of the 3BL, presents a significant challenge in DDARPs. While the literature extensively covers economic aspects, social objectives are less included in the studies. These objectives primarily revolve around optimizing objective functions from the users’ perspective and social welfare. Certain areas of DDARPs, such as hospital problems, tend to place more importance on social objectives. This section summarizes the most common objectives identified in the literature.

6.1. Minimize User Inconvenience and Maximize Service Quality (LOS)

Minimizing user inconvenience and maximizing service quality are essential social objectives in DDARPs because they directly impact passenger interactions and satisfaction with the transportation service. By prioritizing these objectives, transportation providers can ensure that their services are accessible, reliable, and of a high standard, enhancing user satisfaction. Various studies employ different indicators to define and measure the quality of service provided by the system. Achieving these objectives involves addressing multiple sub-objectives, including minimizing passenger waiting times and delivering a high level of service.

Hanne et al. [74] designed a computer-based planning system called Opti-TRANS to solve the intra-hospital transportation problem to maximize service quality, where the service quality was defined as the combination of total lateness and earliness and transportation ride time for patients. Vonolfen et al. [75] used genetic programming to synthesize specialized dispatching rules for the DDARP, where the study focused on minimizing the users’ inconvenience. Raddaoui et al. [9] proposed a system based on the genetic algorithm NSGAII to solve the DDARP with an overall objective of maximizing service quality, where the service quality was measured as a combination of ride time, the number of stations visited and satisfaction of the demand. Issaoui et al. [31] presented a mathematical model to resolve the DAR approach based on NSGAII to improve the quality of service by allocating suitable vehicles to passengers. Lokhandwala and Cai [24] proposed an agent-based simulation model to study the impacts of dynamic ridesharing using traditional and autonomous taxis to minimize inconvenience to the riders. The authors focused on minimizing user inconvenience by minimizing the deviation from the service time. Bian and Liu [76,77] developed a matching optimization program and a Vickrey–Clarke–Groves (VCG) mechanism to determine the optimal vehicle–passenger matching and routing plan to minimize user inconvenience while minimizing operational costs. Xue et al. [78] proposed an improved algorithm framework to study the dynamic ride-sharing service optimization problem, considering passengers’ perceptions of service quality as the objectives, where maximization of passengers’ satisfaction was the main objective. Pfeiffer and Schulz [79] presented an adaptive large neighborhood search (ALNS) to minimize user inconvenience by minimizing the total detour times in a DDARP. Souza et al. [4] proposed a bi-objective mathematical optimization model and a two-stage heuristic for a real-world application of heterogeneous DDARP with no rejects, i.e., a patient transportation system, where the objective was to minimize the users’ inconvenience by estimating shortest paths.

6.2. Minimize User’s Wait Time

User wait time is critical in determining a transportation network’s service quality and performance. It reflects operational efficiency and significantly impacts user satisfaction and perception of service quality [80]. By prioritizing and effectively managing travelers’ wait time, transportation providers can enhance the social sustainability of their operations and contribute to the overall well-being of the users they serve.

Pavone et al. [81] presented decentralized algorithms for stochastic and DVRPs to minimize user wait times. Cortés et al. [82] presented an analytical way of modeling the impact of stochastic rerouting delays for DPDP based on a hybrid adaptive predictive control scheme to optimize vehicle dispatch and routing decisions to minimize user costs based on waiting times. Núñez et al. [80] proposed a multi-objective model-based predictive control approach to solving DPDPs in the context of DDARP to minimize passengers’ wait time while optimizing the total operational cost of the system. Molenbruch et al. [83] considered a generalization of a bi-objective DDARP, incorporating real-life characteristics of patient transportation while minimizing the user’s wait time. Zhang et al. [84] considered the problem of routing a shared fleet of AVs and considered the study’s objective to minimize customer wait times. Alisoltani et al. [85] considered providers’ and passengers’ objectives and proposed an optimization algorithm for vehicle allocation problems, where users’ wait times and travel times were considered the main objectives. Kumar and Khani [86] proposed an optimization model for designing an integrated autonomous mobility-on-demand (AMoD) and urban transit system where the study’s objective was to minimize users’ wait times. Dandl et al. [87] proposed a tri-level mathematical programming model to develop a modeling framework that captures the inter-decision dynamics between mobility service providers and travelers to minimize users’ detours and wait times.

6.3. Optimize Workload Equity

Equity considerations in DDARPs relate to the fair distribution of workloads and the balance of resource utilization. Equitable workload allocation enhances driver acceptance and morale and contributes to the quality of customer service provided [88]. Various studies have employed different metrics to assess workload equity. For instance, Mourgaya and Vanderbeck [89] examined regional compactness and workload balance among routes in period VRPs (PVRPs), where the workload was measured by the total demand served. Similarly, Gulczynski et al. [90] incorporated equity as an objective in PVRPs, considering the number of customers per tour as a workload indicator. Using a weighted sum approach, they integrated workload equity with the total cost objective. One of the most essential DDARP applications with equity aspects focuses on school bus routing. Because the government typically provides these transportation services, equity objectives must be considered in addition to cost efficiency [88]. Parragh et al. [91] formulated a bi-objective model for recurring patient transportation services at the Austrian Red Cross to minimize the average transit time and the total cost. Perugia et al. [92] designed a generalized bi-objective VRPTW of a home-to-work bus service for employees of a large research center in Rome, Italy, with the objectives of optimizing the total cost as well as the quality of the service, measured as the passenger’s time loss compared to direct travel as an overall measure of equity.

6.4. Optimize Equity and Accessibility for Users

In DDARPs, optimizing equity and accessibility for users is crucial to ensure that transportation services are provided in a fair and inclusive manner, regardless of individuals’ demographic or socioeconomic characteristics. Equity optimization ensures that transportation benefits and costs are distributed fairly and appropriately among community members [93]. Wang et al. [94] proposed an equity-oriented VRP to complete the delivery within the time window and further consider the equity of each customer in terms of the earliest possible delivery time for each customer. The authors developed a mono-objective of decreasing the equity cost, which increases delay time beyond the earliest possible delivery time for each customer in the study. Zhang and Khani [95] studied an integrated transit system where ride-hailing services complement the transit service as an efficient and economical access mode, with a trade-off between the objectives, where riders maximize their utilities by deciding whether and how to use the integrated system, while drivers maximize their profits by determining how to serve riders.

By prioritizing accessibility, transportation providers can ensure that their services are available and convenient for all individuals in the community, regardless of their physical abilities or demographic properties. This contributes to the social sustainability of their operations and promotes equal access to opportunities and services. In the context of DDARPs, improving accessibility involves designing vehicle routes and schedules that address users’ specific needs, such as FMLM service, where DAR services are integrated with PT to enhance the overall accessibility of the transportation system. Amor et al. [96] proposed a new formulation of the integrated DARP (iDARP) to improve the users’ accessibility with the objective functions minimizing the total combined costs and waiting time. Perera et al. [97] developed an electric vehicle system to minimize users’ travel time by increasing accessibility to predetermined stops.

7. Environment Dimension Objectives

Recently, the rise of global warming and its associated consequences, including climate change and the high occurrence of natural disasters, have brought attention to the escalating environmental impact. The transportation industry is accountable for 23% of global CO₂ emissions, with road transportation accounting for over 75% of total CO₂ emissions [98,99]. Addressing the environmental dimension has become crucial to fostering sustainable transportation practices that mitigate these challenges. The environmental impact of transportation services can be minimized by optimizing vehicle routes to reduce distances traveled and fuel consumption. By prioritizing sustainable transportation options, providers can contribute to the overall sustainability of their operations while encouraging a more active and healthier lifestyle. These efforts, including reducing vehicle emissions and optimizing fuel efficiency routes, help mitigate transportation’s adverse environmental effects.

Several objectives in the literature prioritize environmental impacts on sustainability, including reducing emissions and minimizing fuel consumption. Economic objectives such as minimizing VDT affect the environment by reducing fuel and energy consumption. The following objectives are commonly found in the literature, focusing on the environmental dimension.

7.1. Minimize Energy and Fuel Consumption

Minimizing energy and fuel consumption is a crucial objective in DDARPs due to their significant environmental and economic benefits. Fuel consumption is based on different criteria, mostly the VDT, the vehicle load, and the type of road taken by the vehicles [100]. By optimizing vehicle routes and prioritizing low-emission vehicles, transportation providers can reduce the amount of energy and fuel consumed while also improving the efficiency of their operations. This objective can benefit the environment and economic benefits for transportation providers by reducing operational costs, improving profitability, and contributing to the overall sustainability of the transportation system [101]. Incorporating this objective into DDARP objective functions can help promote sustainable transportation practices and create a more sustainable and resilient transportation system. Ettazi et al. [100] studied a specific case of VRP in-home health care with the fuel consumption reduction aspect to be optimized.

7.2. Minimize GHG/CO₂/NO_x Emissions

Minimizing greenhouse gas (GHG), carbon dioxide (CO₂), and nitrogen oxide (NO_x) emissions is a crucial objective in DDARPs due to their significant environmental impact. Since the transportation sector is responsible for a large portion of CO₂ emissions, reducing CO₂ emissions while maintaining transportation efficiency has become a more critical issue. By minimizing emissions through optimizing vehicle routes and prioritizing low-emission vehicles, transportation providers can help mitigate the negative environmental impacts of transportation. This objective economically benefits the environment and transportation providers by reducing emissions-related penalties and costs. Incorporating this objective into DDARP objective functions can help promote sustainable transportation practices and contribute to the overall sustainability of the transportation system. Additionally, minimizing emissions can help improve public health and air quality, making it a critical objective from a social perspective. By prioritizing this objective, transportation providers can contribute to a more sustainable, resilient, and socially responsible transportation system.

Kleiner et al. [12] presented a novel dynamic ride-sharing system that minimizes the total emissions by minimizing the total VDT while maximizing the number of served requests. Chevrier et al. [102] proposed a multi-objective DRT problem to minimize emissions by minimizing the number of vehicles used. Atahran et al. [103] proposed a new multi-objective model for the DARP with a fleet of heterogeneous vehicles, with one of the objectives being the minimization of the quantity of CO₂ emitted by the vehicles. Van Heeswijk et al. [104] proposed a planning algorithm for DPDP in intermodal networks, where freight is consolidated by means of reloads to reduce costs and emissions. Chen et al. [98] proposed a bi-objective DDARP to minimize CO₂ emissions and travel time.

7.3. Optimize Traffic Congestion

Optimizing traffic congestion is a crucial aspect of DDARPs, particularly in urban areas with high traffic volumes. Traffic congestion can result in delays, increased fuel consumption, and air pollution, causing negative impacts on the environment and society. Incorporating this objective into DDARP objective functions can help to promote sustainable transportation practices. These objectives benefit transportation providers and the broader society, creating a more sustainable and efficient transportation system that minimizes negative environmental and public health impacts.

The studies considering traffic congestion optimization are solved as a VRP with time-dependent travel times (TDVRP) [105]. There is a variant of VRP known as “City VRP”, which focuses on studies arising in the city, including traffic regulations, traffic congestion, and pollution [106]. Ehmke et al. [107] introduced a planning system for city logistics service providers, which faces those challenges by more realistic vehicle routing considering time-dependent travel times due to congestion. A few studies focus on minimizing energy consumption and CO₂ emissions while reducing traffic congestion. Kok et al. [105] developed a set of DDARP instances on real road networks and a speed model that reflects the critical elements of peak-hour traffic congestion while minimizing the total number of travel routes in congestion. Kim et al. [106] proposed a DDARP model with nonstationary stochastic travel times under traffic congestion. The study aimed to minimize the expected travel cost by visiting all nodes under traffic congestion. Moryadee et al. [108] studied a time-dependent and pollution VRP model, where the objectives were to minimize CO₂ emissions due to traffic congestion. Sabar et al. [109] presented a self-adaptive evolutionary algorithm for the DVRP with traffic congestion. The authors considered various traffic congestion levels in two different routing problems: vehicle routing and traveling salesmen problems. Later, Liu et al. [110] solved a time-dependent VRP with time windows (TDVRPTW) with congestion avoidance. The objective was to minimize overall costs, including costs of hiring the drivers and fuel consumption and emission costs due to congestion.

8. Evaluation and Discussion

Section 8.1 explains the interrelation between the various objectives within the framework of the 3BL dimensions. Subsequently, Section 8.2 analyzes the stakeholders involved with each objective. Section 8.3 further explores the trade-offs among these objectives and examines the interconnectedness or contradictions that may arise between them. Furthermore, Section 8.4 presents a comprehensive review of multi-objective studies conducted in this field, evaluating the interdependencies among objectives from different dimensions in the context of multi-objective problems. Lastly, Section 8.5 presents a geographical analysis, showcasing the continent-specific trends across different dimensions of 3BL.

8.1. Objectives and the Concerned Dimensions

The objectives presented in

Table 1. Summary of the objectives with their concerned dimensions.

Objective	Economic Dimension	Social Dimension	Environment Dimension
Number of vehicles	●
Total costs	●
Total ride time	●
VDT	●
Number of requests	●
LOS	●	●
User’s wait time	●	●
Workload equity	●	●
Equity and accessibility	●	●
Congestion	●	●	●
Fuel consumption	●		●
CO2 emissions	●		●

demonstrate the interplay between these dimensions. For instance, minimizing fuel consumption and emissions addresses both the economic and environmental dimensions by reducing operational costs and promoting environmentally friendly practices. Similarly, objectives focusing on minimizing user inconvenience and maximizing service quality align with the social and economic dimensions by prioritizing user satisfaction and potentially increasing ridership. Minimizing total ride time improves operational efficiency and resource utilization while reducing total VDT, contributing to cost reduction and environmental sustainability. Maximizing total requests served optimizes resource allocation and revenue potential.

Objectives related to workload equity and equity/accessibility for users aim to achieve fairness and inclusivity in the system, benefiting both social and economic aspects. Optimizing traffic congestion encompasses all three dimensions: improving economic efficiency, environmental sustainability, and social welfare.

The interplay between these dimensions is essential for developing comprehensive strategies in transportation systems. Transportation systems can enhance efficiency, user satisfaction, equity, and environmental sustainability by considering economic, social, and environmental factors. This approach promotes cost reduction, resource optimization, user-centric services, and reduced environmental impact. The stakeholders must consider integrating these dimensions into planning and decision-making processes for a balanced and sustainable future.

8.2. Objectives and the Concerned Stakeholders

Transportation decision-making usually involves multiple social interest groups and stakeholders, which are commonly conflicted. Stakeholder roles in the transportation system are defined by their individual needs, objectives, and activities’ private or public nature [111]. In the private stakeholders’ group, we consider service providers, manufacturers, and other private institutions, whereas public actors are environmentalists, policymakers, PT operators, and non-governmental organizations (NGOs). This study identifies and correlates the most critical stakeholders involved in different objectives.

Table 2. Objectives with their concerned stakeholders.

Dimension	Objective	Stakeholders Concerned with the Objective
Economic	Number of vehicles	Service Providers
	Total costs	Service Providers, Manufactures
	Total ride time	Service Providers, Consumers
	VDT	Service Providers, Environmentalists
	Number of requests	Service Providers, Consumers, Policymakers
Social	LOS	Consumers, Policymakers
	User’s wait time	Consumers, Policymakers
	Workload equity	Consumers, Service Providers, Policymakers, PT Operators, Environmentalists, Environmental Organizations
	Equity and accessibility	Consumers, Service Providers, Policymakers, PT Operators, Environmentalists, Environmental Organizations
Environmental	Congestion	Service Providers, Environmentalists, Environmental Organizations, NGOs, PT Operators, Policymakers
	Fuel consumption	Service Providers, Environmentalists, Environmental Organizations, NGOs, PT Operators, Policymakers
	CO2 emissions	Consumers, Service Providers, Policymakers, PT Operators, Environmentalists, Environmental Organizations

contains all the stakeholders concerning the individual objectives of the study. The table presents all the studied objectives in the context of stakeholders concerned with them. The stakeholders involved with each objective include service providers, manufacturers, consumers, policymakers, PT operators, NGOs, environmentalists, and environmental organizations.

Within the economic dimension, stakeholders primarily include service providers and manufacturers. The objective of minimizing the number of vehicles aims to benefit service providers by reducing their fleet size and associated costs. Minimizing operational costs and maximizing profit directly concerns service providers and manufacturers as they aim to optimize their financial performance. Additionally, stakeholders such as environmentalists are concerned with minimizing total VDT to reduce the environmental impact associated with transportation activities.

In the social dimension, minimizing user inconvenience and maximizing service quality involve consumers and policymakers. Consumers seek seamless and comfortable travel experiences, while policymakers aim to ensure high-quality transportation services that meet the needs of the public. Minimizing user wait times also benefits both consumers and policymakers as it enhances the overall efficiency and attractiveness of the transportation system. Workload equity optimization and equity/accessibility for users involve multiple stakeholders, including consumers, service providers, policymakers, PT operators, environmentalists, and environmental organizations. These objectives aim to ensure fairness, equal access, and inclusivity within the transportation system, addressing social equity concerns.

Within the environmental dimension, stakeholders such as service providers, environmentalists, environmental organizations, NGOs, PT operators, and policymakers are concerned with minimizing fuel consumption and emissions. These objectives contribute to sustainable practices, reduced carbon footprint, and improved air quality. The objective of optimizing traffic congestion engages a diverse group of stakeholders, including consumers, service providers, policymakers, PT operators, environmentalists, and environmental organizations. By addressing emissions in the context of traffic congestion, this objective aims to improve environmental sustainability while considering the efficiency and effectiveness of the transportation system.

Understanding the involvement of various stakeholders in achieving these objectives is vital for fostering collaboration, stakeholder engagement, and effective decision-making processes. By considering the perspectives and interests of the concerned stakeholders, transportation systems can develop comprehensive strategies that address economic viability, social welfare, and environmental sustainability.

8.3. Relationships among Various Objectives

Table 3. Objectives’ relationship with each other.

Objectives	Fleet	Total Costs	Ride Time	VDT	No. of Requests	LOS	Wait Time	Workload Equity	Equity and Accessibility	Fuel	CO2	Congestion
Fleet	●	✔	✖	✔	✖	✖	✖	✖	✖	✔	✔	✔
Total Costs	✔	●	✔	✔	✖	✔	✔	✖	✖	✔	✔	✖
Ride Time	✖	✔	●	✔	✖	✔	✔	✖	✔	✔	✔	✔
VDT	✔	✔	✔	●	✖	✖	✖	✖	✔	✔	✔	✔
No. of Requests	✖	✖	✖	✖	●	✔	✔	✔	✔	✖	✖	✖
LOS	✖	✔	✔	✖	✔	●	✔	✔	✔	✖	✖	✖
Wait Time	✖	✔	✔	✖	✔	✔	●	✔	✔	✖	✖	✖
Workload Equity	✖	✖	✖	✖	✔	✔	✔	●	✔	✖	✖	✔
Equity and Accessibility	✖	✖	✔	✔	✔	✔	✔	✔	●	✖	✖	✔
Fuel	✔	✔	✔	✔	✖	✖	✖	✖	✖	●	✔	✖
CO2	✔	✔	✔	✔	✖	✖	✖	✖	✖	✔	●	✔
Congestion	✔	✖	✔	✔	✖	✖	✖	✔	✔	✖	✔	●

shows the interconnections and contradictions among different objectives identified in this paper. The table explains the direct impact of each objective on the others. The interconnections between different objectives are shown as a checkmark (✔) in the table, and the contradictions between different objectives are shown as a cross (✖). It is important to note that the table offers a general overview of the subject matter, and its applicability in real-world situations may involve additional details and complexities in the interplay of the objectives outlined. Moreover, the views of different stakeholders may vary depending on their priorities and perspectives, influencing their perceptions of these interrelationships.

Minimizing the number of vehicles is connected to minimizing fuel consumption and CO₂ emissions, as fewer vehicles result in lower fuel consumption and emissions. However, fewer vehicles can lead to longer pick-up trips due to lower availability, potentially increasing VDT and ride time. Objectives such as minimizing the number of vehicles, minimizing total operational costs, and maximizing total profit are all interconnected objectives primarily concerned with the interests of service providers and manufacturers. Yet, minimizing total costs does not necessarily align with serving as many requests as possible; the most cost-effective solution might involve minimizing operations altogether, conflicting with maximizing service.

Minimizing total ride time, maximizing total requests served, and optimizing equity and accessibility for users are all interconnected objectives that are primarily concerned with the interests of consumers and policymakers. These objectives aim to enhance user satisfaction and ensure fair service access, but they often require higher operational costs and more complex routing strategies. Workload equity, user equity, and accessibility are interconnected objectives that ensure fair and equal access to transportation services for all users.

Minimizing users’ ride time may contradict minimizing total VDT because reducing the distance traveled may increase some users’ time in the vehicle. Minimizing users’ wait time and optimizing workload equity are contradictory objectives because prioritizing workload equity could lead to longer wait times for some users. For instance, ensuring equal workload distribution among drivers may require routing that increases wait times for some passengers, highlighting a trade-off between operational fairness and user convenience. Similarly, optimizing equity and accessibility for users may contradict minimizing total operational cost and maximizing total profit because providing more equitable and accessible services may increase operational costs. For instance, providing equitable and accessible service may necessitate deploying more vehicles or running longer routes, which can increase fuel consumption and CO₂ emissions, as well as operational costs. Minimizing total ride time and minimizing user inconvenience while maximizing service quality are contradictory objectives because some factors that could improve service quality could also increase ride time and inconvenience.

Optimizing emissions due to traffic congestion and minimizing total ride time are contradictory objectives because reducing ride time may require driving at higher speeds, which could increase emissions. This trade-off illustrates the challenge of balancing environmental sustainability with the need for efficient and quick service. Minimizing fuel consumption and minimizing total operational costs may be contradictory objectives because using alternative fuels or more fuel-efficient vehicles may require higher upfront costs for service providers. Minimizing CO₂ emissions and maximizing total requests served may be contradictory objectives because increasing the number of trips could lead to more emissions overall. Balancing these conflicting objectives requires careful planning and innovative solutions, such as integrating electric vehicles, which, while initially costly, can reduce long-term emissions and operational costs.

Improving LOS often leads to better user satisfaction but at the cost of increased operational complexity and costs. For example, enhancing LOS can involve reducing wait times and ride times, which may require additional vehicles and higher fuel consumption, thereby increasing total costs and potentially CO₂ emissions. On the other hand, maintaining high levels of LOS is crucial for attracting and retaining users, indirectly supporting objectives like maximizing the number of requests served and improving equity and accessibility.

Fuel consumption and CO₂ emissions are closely linked with operational and environmental objectives. These environmental objectives can sometimes conflict with maintaining cost-effective operations, particularly when transitioning to more sustainable but expensive technologies. While reducing the fleet size and minimizing VDT contribute to lower fuel consumption and emissions, these measures can adversely affect ride time, wait time, and LOS. Moreover, strategies to reduce congestion, such as optimizing routing and scheduling, can further influence fuel consumption and emissions, highlighting the need for an integrated approach to managing these objectives. Reducing congestion through efficient routing and scheduling can improve ride time and fuel efficiency, reducing emissions. However, measures to reduce congestion, like deploying additional vehicles or optimizing routes, may increase total operational costs and affect the balance of workload equity. Thus, while aiming to reduce congestion, transportation planners must consider the broader implications on cost, equity, and environmental objectives.

In summary, Table 3 discusses how various objectives are interconnected and how they impact each other. Considering these interconnections and contradictions is crucial when designing and implementing transportation systems. The detailed relationships highlighted in the table demonstrate that achieving a single objective in isolation can inadvertently affect others, emphasizing the need for comprehensive planning and innovative solutions to effectively manage the complexities of DDARPs. Therefore, it is essential to consider the interconnections and contradictions between different objectives when designing and implementing transportation systems and to find a balance between competing objectives to achieve an optimal solution.

8.4. Interrelationships among the Objectives in Multi-Objective Problems

The multi-objective optimization approach balances and uses multiple conflicting or interconnected objectives. This study identifies 16 research papers that proposed different multi-objective optimization models to solve the DDARPs. A comprehensive summary of these multi-objective DDARPs used in this review article is provided in Table 4. This table includes multiple studies focusing on multi-objective DDARPs, where the objectives belong to the same or different dimensions of 3BL. Different multi-objective optimization approaches, such as weighted sum, Pareto frontiers, and lexicographic approaches, have been used in these studies.

Table 4 shows that although the objectives in economic and social dimensions are contradictory to each other, there are a few studies that implement them as interconnected. Most studies focus on minimizing inconvenience and total costs [4,70,74,112,113]. Jung et al. [71] focused on multiple objectives, including ride time and total costs, from solely the economic dimension. Other studies [80,114,115] also concentrate on economic and social dimensions, minimizing the total costs and wait time. Raddaoui et al. [9] focus on minimizing the total ride time with the total costs, using the study’s economic and social dimensions.

Regarding objectives from economic and environmental dimensions, studies [104,116] have focused on minimizing total emissions while minimizing total costs. Other studies focus on other objectives from the economic dimension, as Kleiner et al. [12] minimized the total number of served requests with total emissions. Chen et al. [98] focused on including the social dimension as minimizing the ride time with total emissions, utilizing a weighted sum approach, and emphasizing the interconnection between the two objectives.

Two of the studies have included all three dimensions in their multi-objective problems. Atahran et al. [103] combine economic, environmental, and social dimensions, minimizing total costs and CO₂ emissions while maximizing user satisfaction through Pareto frontiers. Hu et al. [117] used a Pareto frontiers approach to solve the multi-objective problem of minimizing fuel consumption, total costs, and total wait time.

The studies in the table contribute valuable insights into multi-objective optimization problems across diverse dimensions. The findings underscore the trade-offs, contradictions, and interdependencies among objectives, providing researchers and decision-makers with valuable approaches and recommendations for balancing conflicting objectives. Overall, the findings suggest that multi-objective optimization approaches can help improve the DDARPs’ performance by considering various objectives and making a trade-off between them. By considering the relationships among objectives and adopting appropriate optimization techniques, decision-makers can make informed choices in complex decision-making scenarios.

Table 4. Multi-objective studies reviewed in this article.

S. No.	Source	Dimensions Involved	Primary Objective	Secondary Objectives	Type of M.O. Approach	Relationship among Objectives	Key Findings
1	Melachrinoudis et al. [113]	Economic, Social	Minimizing Inconvenience	Minimizing Total Costs	Weighted Sum	Contradictory	Weighting objectives differently, the study revealed a substantial time saving of 21.84 h, surpassing the minimal inconvenience increase of 0.46 h caused by additional rides and client wait time.
2	Hanne et al. [74]	Economic, Social	Minimizing Inconvenience	Minimizing Total Costs	Weighted Sum	Contradictory	The study used a weighted sum approach, where minimizing user inconvenience was balanced against minimizing operational costs in ambulance trips. Hospital preferences varied based on whether they operated their ambulance fleet or worked with an external provider.
3	Garaix et al. [112]	Economic, Social	Minimizing Total Costs	Maximizing Service Quality	Lexicographic	Contradictory	The study sequentially prioritized minimizing total operational costs and maximizing service quality using a lexicographic optimization approach.
4	Kleiner et al. [12]	Environmental, Economic	Minimizing Total Emission	Maximizing Number of Served Riders	Pareto Frontiers	Contradictory	The study solves a multi-objective dynamic ride-sharing problem, allowing a trade-off between the minimization of emissions due to the total VKT and the overall probability of a successful ride served.
5	Schilde et al. [70]	Economic, Social	Maximizing Service Quality	Minimizing Total Costs	Lexicographic	Contradictory	The study considers three objectives: maximizing service quality and minimizing operating costs (number of vehicles and total route duration). One approach improves service quality by 11.12–25.86%, while the other optimizes secondary objectives.
6	Raddaoui et al. [9]	Economic, Social	User Inconvenience	Minimizing Ride Time	Pareto Frontiers	Interconnected	The study considers economic and service quality objectives, comparing results for different vehicles and request scenarios. Overall, an improvement is shown when compared to their previous study.
7	Kirchler and Wolfler Calvo [114]	Economic, Social	Minimizing Total Costs	Minimize User Wait Time	Weighted Sum	Contradictory	The study applies weighted objectives, including routing costs and user wait time. The authors present the weight values’ impact on solution quality and support adjusting them for different problem instances.
8	Núñez et al. [80]	Economic, Social	Minimize User Wait Time	Minimize Total Cost	Pareto Frontiers	Contradictory	Five λ values (0 to 1) were used as weights, with λ, 1 − λ assigned to primary and secondary objectives. When λ was increased from 0.75 to 1, the user costs doubled. The study suggests a range of 0.2 to 0.7 to balance user costs and minimize operator impact.
9	Atahran et al. [103]	Environmental, Economic, Social	Minimize Total Costs	Minimize CO2 and Maximize User Satisfaction	Pareto Frontiers	Contradictory and Interconnected	The study combines three heuristics, evaluates performance using quality indicators, and identifies the best solution as a combination of 5% H1, 5% H2, and 90% H3 based on non-dominated solutions.
10	Jung et al. [71]	Economic	Minimizing Travel Time	Minimizing Total Costs	Weighted Sum	Interconnected	The study tackles interconnected objectives with different approaches, optimizing each objective in different scenarios rather than simultaneously.
11	van Heeswijk et al. [104]	Environmental, Economic	Minimize Total Costs	Minimize Emission	Weighted Sum	Interconnected	Results show interconnected objectives with 34% and 30% savings in total costs and CO2 emissions, respectively. Adding route flexibility yields an extra 5% and 9% improvement in costs and CO2.
12	Hu et al. [117]	Environmental, Economic, Social	Minimize Fuel Consumption	Minimize Total Costs, Minimize Wait Time	Pareto Frontiers	Contradictory and Interconnected	The study compared single and multi-objective approaches using relative percentages for total costs. Multi-objective model achieved 34%, 21%, and 34% for objectives, with the total cost at 90%, outperforming single objective solutions at 95%, 108%, and 106%.
13	Abedi et al. [116]	Environmental, Economic	Minimize Total Costs	Minimize CO2	Pareto Frontiers	Interconnected	Authors assigned weights and penalties to estimate multiple objectives, plotting focal points between routing costs and CO2 emissions to demonstrate the impact of weight allocation on the Pareto front.
14	Souza et al. [4]	Economic, Social	Minimize Total Costs	Minimize User Inconvenience	Weighted Sum	Contradictory	The study analyzes two scenarios, assigning weights to objectives and solving the problem for various weight combinations. The problem was solved for all the weights from (0.1, 0.9) to (0.9, 0.1). Results demonstrate a trade-off between transportation costs and user inconvenience as weights vary.
15	Chen et al. [98]	Environmental, Social	Minimize Travel Time	Minimize Emission	Weighted Sum	Interconnected	The study finds an interconnection between travel time and emissions. Different vehicle routes emerge for emission and travel time minimization, except in equal-weight bi-objective problems favoring travel time.
16	Nasri et al. [115]	Economic, Social	Minimize Total Costs	Minimize User Wait Time	Lexicographic	Contradictory	The study applies lexicographic order to prioritize total cost and user wait time in multi-objective problem-solving. Results show improved total waiting time regardless of objective priority. Cost prioritization improves 7 out of 12 instances, while wait-time prioritization improves all instances except one.

Table 4. Multi-objective studies reviewed in this article.

S. No.	Source	Dimensions Involved	Primary Objective	Secondary Objectives	Type of M.O. Approach	Relationship among Objectives	Key Findings
1	Melachrinoudis et al. [113]	Economic, Social	Minimizing Inconvenience	Minimizing Total Costs	Weighted Sum	Contradictory	Weighting objectives differently, the study revealed a substantial time saving of 21.84 h, surpassing the minimal inconvenience increase of 0.46 h caused by additional rides and client wait time.
2	Hanne et al. [74]	Economic, Social	Minimizing Inconvenience	Minimizing Total Costs	Weighted Sum	Contradictory	The study used a weighted sum approach, where minimizing user inconvenience was balanced against minimizing operational costs in ambulance trips. Hospital preferences varied based on whether they operated their ambulance fleet or worked with an external provider.
3	Garaix et al. [112]	Economic, Social	Minimizing Total Costs	Maximizing Service Quality	Lexicographic	Contradictory	The study sequentially prioritized minimizing total operational costs and maximizing service quality using a lexicographic optimization approach.
4	Kleiner et al. [12]	Environmental, Economic	Minimizing Total Emission	Maximizing Number of Served Riders	Pareto Frontiers	Contradictory	The study solves a multi-objective dynamic ride-sharing problem, allowing a trade-off between the minimization of emissions due to the total VKT and the overall probability of a successful ride served.
5	Schilde et al. [70]	Economic, Social	Maximizing Service Quality	Minimizing Total Costs	Lexicographic	Contradictory	The study considers three objectives: maximizing service quality and minimizing operating costs (number of vehicles and total route duration). One approach improves service quality by 11.12–25.86%, while the other optimizes secondary objectives.
6	Raddaoui et al. [9]	Economic, Social	User Inconvenience	Minimizing Ride Time	Pareto Frontiers	Interconnected	The study considers economic and service quality objectives, comparing results for different vehicles and request scenarios. Overall, an improvement is shown when compared to their previous study.
7	Kirchler and Wolfler Calvo [114]	Economic, Social	Minimizing Total Costs	Minimize User Wait Time	Weighted Sum	Contradictory	The study applies weighted objectives, including routing costs and user wait time. The authors present the weight values’ impact on solution quality and support adjusting them for different problem instances.
8	Núñez et al. [80]	Economic, Social	Minimize User Wait Time	Minimize Total Cost	Pareto Frontiers	Contradictory	Five λ values (0 to 1) were used as weights, with λ, 1 − λ assigned to primary and secondary objectives. When λ was increased from 0.75 to 1, the user costs doubled. The study suggests a range of 0.2 to 0.7 to balance user costs and minimize operator impact.
9	Atahran et al. [103]	Environmental, Economic, Social	Minimize Total Costs	Minimize CO2 and Maximize User Satisfaction	Pareto Frontiers	Contradictory and Interconnected	The study combines three heuristics, evaluates performance using quality indicators, and identifies the best solution as a combination of 5% H1, 5% H2, and 90% H3 based on non-dominated solutions.
10	Jung et al. [71]	Economic	Minimizing Travel Time	Minimizing Total Costs	Weighted Sum	Interconnected	The study tackles interconnected objectives with different approaches, optimizing each objective in different scenarios rather than simultaneously.
11	van Heeswijk et al. [104]	Environmental, Economic	Minimize Total Costs	Minimize Emission	Weighted Sum	Interconnected	Results show interconnected objectives with 34% and 30% savings in total costs and CO2 emissions, respectively. Adding route flexibility yields an extra 5% and 9% improvement in costs and CO2.
12	Hu et al. [117]	Environmental, Economic, Social	Minimize Fuel Consumption	Minimize Total Costs, Minimize Wait Time	Pareto Frontiers	Contradictory and Interconnected	The study compared single and multi-objective approaches using relative percentages for total costs. Multi-objective model achieved 34%, 21%, and 34% for objectives, with the total cost at 90%, outperforming single objective solutions at 95%, 108%, and 106%.
13	Abedi et al. [116]	Environmental, Economic	Minimize Total Costs	Minimize CO2	Pareto Frontiers	Interconnected	Authors assigned weights and penalties to estimate multiple objectives, plotting focal points between routing costs and CO2 emissions to demonstrate the impact of weight allocation on the Pareto front.
14	Souza et al. [4]	Economic, Social	Minimize Total Costs	Minimize User Inconvenience	Weighted Sum	Contradictory	The study analyzes two scenarios, assigning weights to objectives and solving the problem for various weight combinations. The problem was solved for all the weights from (0.1, 0.9) to (0.9, 0.1). Results demonstrate a trade-off between transportation costs and user inconvenience as weights vary.
15	Chen et al. [98]	Environmental, Social	Minimize Travel Time	Minimize Emission	Weighted Sum	Interconnected	The study finds an interconnection between travel time and emissions. Different vehicle routes emerge for emission and travel time minimization, except in equal-weight bi-objective problems favoring travel time.
16	Nasri et al. [115]	Economic, Social	Minimize Total Costs	Minimize User Wait Time	Lexicographic	Contradictory	The study applies lexicographic order to prioritize total cost and user wait time in multi-objective problem-solving. Results show improved total waiting time regardless of objective priority. Cost prioritization improves 7 out of 12 instances, while wait-time prioritization improves all instances except one.

8.5. Geographic Analysis

This section presents the continent-wise geographical analysis of the papers used in the survey. It shows which continents focus on which type of dimensions in the objectives. This study uses 98 papers for the review; Figure 8 shows the distribution of the number of papers in each dimension. This study considers 55 papers focusing on the economic dimension, 28 on the social dimension, and 15 on the environmental dimension. This shows that the maximum number of papers, i.e., 56.1%, focus on objectives related to the economic dimension. The environmental dimension is the least explored dimension, with 15.3% of papers focusing on environmental objectives. Almost 28.6% of papers consider the social dimension as their primary objective.

Later, Figure 9 shows the distribution of the number of papers in each dimension across different continents. The figure shows that Europe contributes to the maximum number of papers in each dimension. The results show that North America contributes to the second most papers on the economic and social dimensions but has no studies on the environmental dimension. Asia has the second most papers in the environmental dimension while contributing to the few papers in the social dimension.

8.6. Implications

This research provides a comprehensive understanding of the objectives of the DDARPs through the lens of the 3BL approach, encompassing economic, social, and environmental dimensions. The implications of this study are significant for both theoretical advancements and practical applications of DDARPs in transportation management. This study advances the theoretical understanding of how different dimensions of sustainability intersect and influence each other. By highlighting the interconnections and contradictions among various objectives, this research provides a detailed perspective on the complexity of achieving balanced and sustainable outcomes, as explored and discussed in detail in Section 8.3.

Several case studies from major cities illustrate the practical integration of the 3BL approach in transport planning. In New York City, the “Rebuilding New York’s Transportation System” project, part of the MTA’s 2020–2024 Capital Program, invests $51.5 billion to modernize subway and bus networks, reducing operational costs, emissions, and traffic congestion, while generating 350,000 jobs and improving service accessibility [118,119]. Los Angeles integrates electric buses, reducing emissions and operational costs with wireless charging infrastructure, and enhances social equity by providing sustainable transportation in low-income areas. This case study demonstrates the practical application of multi-objective optimization where economic efficiency, environmental sustainability, and social equity are considered simultaneously [120]. Copenhagen’s comprehensive approach includes extensive cycling infrastructure and efficient PT systems that reduce vehicle emissions and traffic congestion. By promoting cycling and PT, Copenhagen has lowered travel costs for residents and enhanced accessibility, which contributes to social equity. The environmental benefits of reduced emissions and improved air quality also lead to better public health outcomes, further demonstrating the city’s success in balancing the three dimensions of sustainability [121]. These examples reinforce the importance of adopting a comprehensive approach to transportation management, recognizing and leveraging the interplay between economic, social, and environmental factors to create sustainable and equitable urban mobility solutions.

Building on these insights, this study provides recommendations for policymakers, service providers, and governmental agencies. It underscores the importance of involving various stakeholders in transportation planning and policy development, as explained in Section 8.2. By leveraging the 3BL framework, policies can be designed to promote sustainable mobility that addresses the diverse needs of stakeholders. Policymakers should consider the interconnectedness of economic, social, and environmental objectives to develop comprehensive transportation initiatives. Transportation service providers can use multi-objective optimization to develop strategies that minimize costs and emissions while maximizing service quality and accessibility. The government can advocate for equitable transportation options, and the consumers can provide valuable feedback and support for sustainable initiatives. By recognizing and addressing the complex interdependencies among these dimensions, more sustainable and equitable transportation systems can be achieved for future urban mobility.

9. Conclusions and Future Remarks

The DDARPs are one of the most critical operations in solving the real-life problems of pickup and drop-off. The recent literature on DDARPs considers more real-life problems with different objectives in relation to the solution approaches. Considering this, it is essential to understand the implication of the different objectives used in the studies on the sustainability aspect of the transportation system.

Considering the need to align DDARP objectives with sustainability priorities, this study classifies different objectives of the reviewed studies under the three dimensions of the 3BL approach. The 3BL implements and evaluates the principles of sustainable development and allows for progress to be achieved in multiple dimensions, often identified as the environment, the economy, and society. This paper has reviewed and classified the relevant recent literature on the objective functions of the DDARPs. While existing review articles focus on the solution approaches, this study focuses on the objectives used in solving the problems. To the best of our knowledge, this is the first conceptual review focusing on the main objectives of DDARPs.

This study analyzes the trade-offs, interconnections, and contradictions that may arise among the studied objectives. The results show that the social and environmental dimensions also directly impact the economic dimension, as all the objectives affect the operational costs. Similarly, objectives focusing on the economic dimension generally contradict social dimension objectives. In addition, this study also provides a comprehensive review of 16 multi-objective studies used in this study, evaluating the interdependencies among objectives from different dimensions. The findings indicate that different approaches, including weighted sum, Pareto frontiers, and lexicographic approaches, have addressed multi-objective problems in transportation systems. The studies demonstrate the interconnectedness and interdependencies among objectives from different dimensions.

This study identifies critical stakeholders in each objective, highlighting the importance of considering their perspectives and interests for effective decision-making. Private stakeholders, including service providers and manufacturers, are primarily concerned with economic objectives. At the same time, public actors such as policymakers, environmentalists, and NGOs are involved in social and environmental objectives. The results from the geographical analysis indicate that the maximum number of studies focus on the economic dimension of 3BL, followed by social and environmental dimensions. European countries contribute the maximum number of papers in each dimension. North America focuses mainly on economic and social objectives and has no studies conducted in the environmental dimension. Asia produces the third most number of papers in overall DDARPs and second most in the environmental dimension.

There are a few limitations to the present study. These limitations are significant to note as they may have impacted the overall comprehensiveness of the review. Firstly, the review only included studies published in English, while studies published in other languages were excluded. Additionally, publication selection bias may have occurred, resulting in the exclusion of some relevant articles. This study aimed to develop a conceptual review of the objectives of DDARPs, focusing on classifying them within the 3BL approach to sustainability. However, the selection of review articles was limited and may not have provided a comprehensive review of the literature on the objectives of DDARPs.

However, it is worth noting that Table 1, Table 2 and Table 3 only provide a high-level overview, and there may be more nuance or complexity in how these factors interact with one another in real-world scenarios. Additionally, stakeholders may have different priorities or perspectives that could impact how they view the relationships between these factors.

This study is a first step toward characterizing and evaluating the trends of DDARPs toward creating a sustainable transportation system. Future studies can focus on explaining all the application areas of DDARPs and present literature reviews on the topic. This study’s outcome is a path for future studies to work toward the less explored objectives. This study suggests the requirements for solving DDARPs with social and environmental objectives. The papers published in North America can focus on more studies with objectives focusing on the environmental dimension. There is a gap in the literature on studies focusing on equity, i.e., more studies should focus on multi-objective problems considering several objectives across all dimensions and can optimize equity.