Sustainable Mobility as a Service: A Scientometric Review in the Context of Agenda 2030

Abstract

City planners are facing an emergency to develop, promote, and implement actions that allow the mobility needs of people and businesses in cities (and their surroundings) to be satisfied to assure a better quality of life. Among the different actions to promote, there is public transport, which should be the cornerstone of sustainable urban mobility. The only way to achieve the transition from private to public is by integrating services into a multimodal network and then encouraging interchange between different modes of transportation. In this context, the development of mobility as a service (MaaS) solutions is significant, and different studies have been developed in recent years. The paper thus introduces a scientometric review of such a topic in relation to the Sustainable Development Goals of Agenda 2030. The study focuses on the current state of MaaS implementation, trends, and research gaps, with an in-depth analysis of emerging themes, based on 819 documents selected from the WoS and Scopus databases. Introducing the database, and research methodology, an accurate interpretation of the data generated by the bibliometric analysis, and the primary evaluation parameters are outlined. The identified studies have been then categorized into three thematic groups with the intention of offering a comprehensive study that identifies the shortcomings and difficulties in the research carried out in these areas up to today. Particular attention is paid to how this research relates to the Sustainable Development Goals.

1. Introduction

Mobility as a service (MaaS) is an emerging and evolving phenomenon, representing a technology-driven innovation that enables users to plan and undertake their journeys through the seamless integration of various mobility modes offered by different service providers, all within a single IT-based interface [1,2]. Beyond its technological novelty, MaaS has the potential to serve as a powerful catalyst for sustainable urban development, particularly within the framework of the 2030 Agenda of the United Nations, which places mobility and transport as the central pillars of sustainable development [3].

Mobility is a positive decisive factor for the social and economic development of urban structures, as mobility can have a negative factor in environmental impacts, under specific conditions and given cases [4,5,6]. Transport is responsible for almost 30% of carbon dioxide emissions in the European Union, 72% of which comes from road transport [7]. The adoption of MaaS as a smart mobility solution presents a unique opportunity to mitigate environmental impacts by promoting the use of more sustainable transportation options, leading to the creation of more efficient and effective transportation systems through reduced consumption, lower emissions, and increased access to services and data [8]. In addition to its environmental benefits [9,10,11], MaaS is essential to driving social sustainability in urban areas. According to the UN [12], effective mobility strategies must go beyond reducing emissions to promote social equity, inclusion, and accessibility [13], ensuring that vulnerable populations gain improved access to essential services, fostering social cohesion, and improving quality of life in urban areas.

In this dynamic context, Italy has launched the “Mobility as a Service for Italy” (MaaS4Italy) initiative, funded through the National Recovery and Resilience Plan under the NextGenerationEU program. This project aims to create a unified digital platform that integrates various transport services, improving accessibility, multimodality, and sustainability in Italy. Originally built on previous successful trials in Naples, Rome, Florence, Milan, and Turin, the initiative will now be extended to seven new regions [14].

Despite this stimulating context and the various advances in the development of the MaaS concept, its concrete development is still limited [15]. According to Vitetta [16], the MaaS concept should evolve towards sustainable MaaS (S-MaaS). The initial phase, often referred to as MaaS 1.0 or I-MaaS (ICT MaaS), introduces the integration of services through an information and communications technology platform, offering users digital services such as information, booking, and payment within an integrated transport system. As digitalization and ICT evolution processes improved [17], the MaaS framework advanced to what is often termed MaaS 2.0 or T-MaaS (TSM and ICT MaaS). It has been built on the foundations of MaaS 1.0 by incorporating transport system management (TSM) and decision support systems (DSS), thereby determining a more proactive approach to designing and managing transport supply and demand [2]. The current research trend and the focus of future developments are on MaaS 3.0 or S-MaaS (sustainable MaaS), which represents a significant evolution within the MaaS framework by integrating sustainability as a fundamental component. In particular, again, the MaaS 3.2 is distinguished by its alignment of transport systems with the Sustainable Development Goals (SDGs) outlined in Agenda 2030, with several studies beginning to analyze the demand side [18], the supply side [19], and the demand–supply interaction [4].

Many literature reviews have been conducted to integrate the efforts and experience of MaaS development. Many case study-based reviews offered insight into the practical challenges and successes of MaaS implementations in various contexts, such as Randstad [20], South Africa [21], Global South [22], Zagreb [23], and Hong Kong [24], with [25] focusing on the Moovit mobility app in the United States. Among reviews focused on specific themes, Cisterna et al. and Turno et al. [26,27,28] focused on acceptance of MaaS and user preferences, while Daniela et al. [29] analyzed 21 MaaS-like schemes around the world. Another important theme is the impact of MaaS on the transportation and automotive industry, as explored in [30], and Todorovic et al. [31] further investigated the role of electric vehicles (EVs). Arias-Molinares et al. and Maas [32,33] investigated general publication resources and main topics of MaaS implementation, analyzing 57 and 127 articles, respectively. Regarding sustainability, Sakai [34] explored MaaS trends and policy-level initiatives within the EU, while Pritchard [35] addresses sustainability more broadly, encompassing environmental and social dimensions, although it does not explicitly refer to Agenda 2030. Existing reviews have contributed significantly to the presentation of the research history of MaaS and to the expansion of its future horizons; however, efforts to employ objective and quantitative methods to map the visualized relationships among researcher clusters, collaborative nations, and high-frequency keywords in this field remain limited, as are efforts to align MaaS development with the SDGs.

To address these gaps, this study aims to develop a systematic review utilizing scientometric techniques to explore the current status, trends, focus, and gaps in MaaS research in the context of Agenda 2030. The contributions of this study, which differ from previous work [20,24,34,35,36,37,38], are summarized as follows:

• studies from multiple academic databases based on the significant academic progress in MaaS research in recent years are retrieved, selected, and systematically analyzed, aiming to identify the trends of MaaS research, particularly in the context of Agenda 2030;
• the scientometrics method is implemented in order to quantitatively map bibliographical networks of authorship, countries, and keywords co-occurrence for MaaS literature to avoid subjectivity and arbitrariness;
• the hotspots and academic frontiers of MaaS research need to be categorized and discussed more in depth; such an identification should help to identify any knowledge gaps to guide future studies towards sustainability, including the transition to mobility as a feature (MaaF).

This paper is organized as follows. After this introduction, Section 2 describes the adopted methodology, and Section 3 presents the results of the scientometric analysis, which is concluded in Section 4 with the discussion of the main identified research topics. Finally, in Section 5, conclusions are presented, offering a summary of the identified gaps in MaaS implementation and recommendations for future studies.

2. Materials and Methods

A systematic review is a rigorous form of literature review that applies a structured and methodical approach to identifying, synthesizing, and evaluating scientific evidence within a specific field. This process is carried out with strict adherence to predetermined criteria for the eligibility and relevance of studies, ensuring a comprehensive search to capture all qualifying research, and is designed to be transparent and reproducible, guaranteeing the reliability of the findings [39]. Scientometric analysis is a quantitative tactic that could enhance the visual and logical perception of the results of systematic review by assessing, grouping, and mapping the quality and relevance of articles through mathematical models and algorithms [40].

The three-phase process of this study is illustrated in Figure 1. The research framework is organized as follows:

• in the first phase (literature search and selection), a comprehensive search on MaaS is performed within the main scientific database (i.e., Scopus and WoS); Scopus is recognized as the largest single abstract and indexing database published, and Web of Science (WoS) has shown to have some key advantages, such as its depth of coverage [41]; these databases complement each other effectively, as neither resource is entirely comprehensive on its own; subsequently, two searches have been merged and the duplicates have been removed by EndNote 21©; then, the definition of exclusion criteria is defined and implemented;
• in the second phase (scientometric analysis), the visualization of the bibliographic network and quantitative analysis are developed using one of the most powerful softwares (i.e., CiteSpace 6.3.R3 Advanced©; [40,42]); in this phase, the analysis focuses on

  ○

  analysis of publication trends, in order to capture the evolution of research output over time;

  ○

  collaboration network among nations analysis, in order to investigate the global distribution of cooperative research efforts and the patterns of collaboration between different countries in the selected studies;

  author collaboration network analysis, examining the relationships among authors and the structure of their collaborative networks within the research community;

  ○

  keywords analysis and clustering, in order to identify key themes and emerging research trends over different time spans;

• in the third stage, three themes are then discussed separately based on the clustering results; finally, the gaps and research findings are summarized.

To ensure an accurate interpretation of the data generated by the bibliometric analysis, the primary evaluation parameters are outlined below.

• the burst strength (BS) is a metric that quantifies citation bursts, which refers to the specific time window during which an author or a study experiences a sudden increase in citations [43];
• two parameters, silhouette (S) and modularity (Q), are used to measure the overall structural properties of the network; a higher S value indicates a better degree of matching between this node and its generic clustering, while a higher Q value denotes a higher degree of dispersion of the network [44];
• the betweenness centrality (w) is a key metric used to assess the importance of nodes within a network; this metric quantifies the extent to which a given node serves as an intermediary between other pairs of nodes within the network [45]; the centrality of node i can be defined according to Equation (1):

$${\omega }_i={\sum }_{m\ne i\ne k} {\displaystyle \frac{{\mu }_{mk}\left(i\right)}{{\mu }_{mk}}}$$

(1)

where

○

${\mu }_{mk}$ is the sum of shortest paths from node m to node k,

○

${\mu }_{mk}\left(i\right)$ is the number of those paths that pass-through node i.

3. Results

This section discusses the application of the previously presented methodology, with specific reference to phases 1 and 2. The results will enable categorization into three thematic groups, which will help highlight emerging topics in MaaS.

3.1. Literature Search and Selection

Based on the literature analysis aligned with the objectives of this study, several recurring keywords were identified: “MaaS”, “Mobility as a Service”, “Mobility-as-a-Service”, “MaaS systems”, “MaaF”, “Mobility as a Feature”, “Mobility-as-a-Feature”. The keywords analyzed include two sources: (i) original keywords offered by authors in their articles, and (ii) extended keywords based on the subject classification of a journal or database [46].

Results of the overall search are shown in

Table 1. Overall search results.

Logical Statement	“MaaS” OR “Mobility as a Service” OR “Mobility-as-a-Service” OR “MaaS Systems” OR “MaaF” OR “Mobility as a Feature” OR “Mobility-as-a-Feature”
Database	Number of Papers	Papers Net of Duplicates
WoS	783	1171
Scopus	1094

; the performed search in WoS and Scopus databases using the OR operator, applying filters to include only English-language papers published up to the first semester of 2024, yielded respectively 783 (WoS) and 1094 (Scopus) papers. In total, 1877 papers have been selected. These papers were then imported into the software EndNote 21© to facilitate the merging of the two libraries and the automatic and manual removal of duplicates. Then, 1171 studies have been identified.

In the third stage of the selection process (phase 1), a manual review of the titles and abstracts (and, if necessary, the full texts) of each of the 1171 articles has been carried out to exclude publications not strictly relevant to the MaaS domain. The primary exclusion criteria, listed in

Table 2. Pre-specified exclusion criteria.

Articles Excluded	Exclusion Criteria
208	Articles not relevant to the transport sector
125	Articles associated with other MaaS/MaaF acronyms
19	Transport-related articles without a MaaS connection

, were as follows: articles not relevant to the transport sector, leading to the exclusion of 208 articles; articles associated with other MaaS/MaaF acronyms, resulting in the exclusion of 125 articles (46.4% of which referred to the Mycosporine-Like Amino Acids and 29.7% of which referred to the Mindful Attention Awareness Scale); and transport-related articles without MaaS connection, leading to the exclusion of 19 articles.

At the end of this analysis, 819 articles were confirmed as relevant. To maintain comprehensive records of all bibliometric data, the library was exported from EndNote 21© in .ris format to facilitate seamless integration with CiteSpace 6.3.R3 Advanced©.

3.2. Scientometric Analysis

The annual trend of the 819 selected articles is illustrated in Figure 2. Since 2018, the number of published articles has shown a significant year-over-year increase, with the peak observed in 2023 at 168 publications, representing 95% (779 out of 819) of the total publications.

To better understand the publication trends related to MaaS, a quadratic regression model was applied to the data from 2014 to 2023. A one-variable polynomial regression model, considering the observations n = 10 and a quadratic regression, can be expressed through Equation (2) [47]:

$$y_i={\beta }_0+{\beta }_1{x}_i+{\beta }_2{x}_i^2+e_i \mathrm{for} i=1, 2,\dots ,n$$

(2)

where

• $y_i$ is the value of the dependent variable Y for the i-th case (i.e., the number of papers published in the i-th year),
• $x_i$ is the value of the dependent variable X for the i-th case,
• $e_i$ is the random error component for the i-th case,
• β₀, β₁ and β₂ are the regression coefficients to be estimated.

The generalized least squares (GLS) method is used for parameter estimation. The mean squared error (MSE) is an unbiased estimator of the variance ${\sigma }^2$ of the random error term ($e_i$) and is defined as follows [47]:

$$MSE={\displaystyle \frac{\sum _{i=1}^n {\left(y_i-{\widehat{y}}_i\right)}^2}{n-(k+1)}}$$

(3)

where

• $y_i$ are the observed values,
• ${\widehat{y}}_i$ are the fitted values of the dependent variable Y for the i-th case.

Since the mean squared error is the average squared error, where averaging is done by dividing by the degrees of freedom, the MSE is a measure of how well the regression fits the data. The root means squared error $RMSE (=\sqrt{MSE}$) is an estimator of the standard deviation σ of the random error term. For this analysis, the MSE has been calculated to be 132.64, and the RMSE was found to be 11.52.

Based on the projections derived, where the number of published papers is expected to reach 201 in 2024, the overall trend indicates that despite being an emerging phenomenon, MaaS is gaining increasing importance in the research community.

Data related to the 819 selected papers have been mapped, and the exploration and visualization of the overall content and temporal evolution of research within a specific field have been performed. The analyses focused on collaborations among nations, authors, and keywords. These analyses were carried out setting the selection criteria to the index g index with a parameter of k = 25, applying a one-year time slice. To avoid reductions in the size of the generated networks, no pruning techniques have been applied.

3.2.1. Collaboration Network

The geographic analysis allows the level of collaboration between research institutions of different countries to be clearly highlighted. A network consisting of 69 nodes and 218 edges has been generated, as shown in Figure 3. The larger nodes represent countries with a high number of studies, while the different colors of the edges indicate collaborative relationships over various years of publication (from 2014 to 2024). Countries such as Germany (centrality = 0.35), the UK (centrality = 0.36), and Australia (centrality = 0.11) are highlighted by purple circles, indicating their role as key hubs in international collaboration in MaaS research, with a betweenness centrality value greater than 0.1.

In

Table 3. Top 6 most productive countries.

No.	Country	Frequency	Centrality
1	Germany	104	0.35
2	Australia	78	0.11
3	United Kingdom	76	0.36
4	Italy	65	0.04
5	Japan	62	0.07
6	United States	54	0.06

, the six most productive countries are listed, ranked by the number of published contributions, along with their values of centrality of interdependence within the network to assess their role in collaboration with other nations. These countries are Germany (104 articles, centrality = 0.35), Australia (78 articles, centrality = 0.11), UK (76 articles, centrality = 0.36), Italy (65 articles, centrality = 0.04), Japan (62 articles, centrality = 0.07), and US (54 articles, centrality = 0.06).

The substantial presence of European countries in the data aligns with significant European experience in the field of MaaS. In particular, in 2017, Helsinki, the capital of Finland, was the site of the first commercial MaaS solution, known as Whim, introduced by the private operator MaaS Global [48]. These data were converted to keyhole markup language (KML) and imported into Google Earth© for real-world visualization. As shown in Figure 4, when considering the entire period from 2016 to 2024 (a), the results are consistent with the data plotted. However, focusing on the period 2016–2018 (b), it becomes evident that Europe has been at the forefront of MaaS adoption since its inception.

Nevertheless, what is particularly noteworthy is the significant contribution of Australia and the United States to the MaaS landscape, despite the fact that both countries historically have shown a high reliance on private transport. For example, in Australian cities, investment in private car infrastructure dwarfs investment in public transport investment; in Sydney, annual toll costs often exceed fuel costs [49]. However, in Australia and other cultures dependent on cars, such as the United States, there appears to be a general decline in private car use [50], which presents a favorable context for the growth and adoption of MaaS services.

As shown above, MaaS is receiving more and more attention from academics and policymakers, and the present understanding of MaaS development is mainly based on studies that have examined pilot projects, mostly carried out in North America and Europe, as well as in Australia. In contrast, as reviewed by Chen and Acheampong [51] by examining the development of MaaS in mainland China, significant development has also been reached in China with the relevant active MaaS platforms and projects throughout Chinese cities. However, further effort needs to be done in order to push the progress from experimental programs to widespread adoption to date [52]. Currently, MaaS platforms are either implemented as standalone apps or are integrated into pre-existing social networking and navigation systems, and they only offer limited information and modal/service connectivity.

Two primary MaaS delivery models were identified: the partnership model, which is gaining popularity in cities like Beijing to integrate more mobility services beyond traditional public transportation, and the public-controlled model, which is more common and focuses primarily on integrating already-existing state-run public transport services. Then, although MaaS is still a relatively recent concept, it is steadily rising to the top of the priority list in national and subnational governmental policies and actions, and it is viewed as essential to achieving China’s long-standing multimodal transport integration imperatives. As a result, mainland China policies and governments are supporting efforts to spread the presence of MaaS.

3.2.2. Authorship Collaboration Network

Using “author” as the node type, a network comprising 311 nodes and 281 edges, shown in Figure 5, has been built. To make the network more compact and display valuable information, the node settings have been configured to show only the largest k = 1 connected component. According to the data presented in

Table 4. Top 6 most productive authors.

No.	Scholar	Affiliation	Frequency
1	David Hensher	University of Sydney (AU)	18
2	Corinne Mulley	University of Sydney (AU)	11
3	Göran Smith	University of Sydney (AU)	10
4	Jana Sochor	Chalmers University of Technology (SE)	8
5	Athena Tsirimpa	The American College of Greece (GR)	8
6	Amalia Polydoropoulou	University of the Aegean (GR)	8

and consistent with the results of the geographic analysis, the three most productive authors are affiliated with the University of Sydney (AU). These are David Hensher (18 papers), Corinne Mulley (11 papers), and Göran Smith (10 papers).

By analyzing the connections of these three key nodes, the most active research community in the MaaS field has been identified, in order to highlight the driving force in advancing research and development within the MaaS domain. This community includes prominent collaborators such as Chinh Q. Ho, John D. Nelson, Daniel J. Reck, Chinh Ho, Jana Sochor, I.C. MariAnne Karlsson, Hans Arby, Steven Sarasini, and Helena Strömberg, who frequently collaborate with the above most productive authors, i.e., David Hensher, Corinne Mulley, and Göran Smith.

3.2.3. Keyword Analysis and Clustering

A keyword is a noun or phrase that not only reflects the meaning or core content of a paper but also reveals the long-term development of a specific research field [53]. Therefore, the keyword co-occurrence analysis can dig out the current research hotspots of MaaS. As said above, the keywords used include the original keywords and the extended keywords based on the subject classification.

To point out and assess the topics that gained the most attention and emerging trends within the field of MaaS, a citation burst analysis was carried out to identify keywords of particular interest to the scientific community during specific periods. Figure 6 presents the six keywords that exhibited a significant increase in academic interest, as indicated by their burst strength (BS) values, over a defined “burn start” and ‘burn end’ period. In particular, the keywords “numerical model” and “willingness to pay” achieved the highest BS values (4.13 and 4.10, respectively) between 2020 and 2021. These are followed by the keywords “transport systems” and “public transit” (BS = 3.68, 2021–2022), “electric vehicles” (BS = 3.06, 2017–2019), and “mobility on demand” (BS = 2.76, 2019–2021).

Cluster analysis is used to record and classify specific phenomena systematically and quantitatively by grouping elements that share common characteristics or attributes [54]. In this study, a hard clustering analysis of keyword co-occurrence, based on the latent semantic indexing (LSI) metric, has been used to identify meaningful research topics. The application of LSI, an advanced textual analysis technique that uncovers the underlying semantic relationships between terms within a data corpus [55], provided an effective and comprehensive representation of these relationships.

Figure 7 presents the network map of 11 noticeable clusters. They are: #0 “Latent class cluster analysis” (31 elements, S = 0.718), #1 “User perspective” (27 elements, S = 0.724), #2 “Sustainable mobility” (25 elements, S = 0.775), #3 “Shared mobility” (28 elements, S = 0.764), #4 “Autonomous vehicle” (32 elements, S = 0.761), #5 “Urban mobility digitalization” (23 elements, S = 0.775), #6 “Literature review” (17 elements, S = 0.859), #7 “Risky choice” (12 elements, S = 0.867), #8 “Neural network approach” (10 elements, S = 0.936), #9 “New technologies” (2 elements, S = 0.996), and #10 “Single-leader multi-follower game” (3 elements, S = 0.997). The values of modularity and weighted mean silhouette of the network are 0.4902 and 0.7865, respectively.

Figure 8 presents a temporal view of the 11 identified groups, providing an overview of the evolution of the topics covered between 2014 and 2024. This visualization includes the predominant keywords associated with each cluster, highlights the main topics in each research area, and illustrates the connections both within and between different clusters. The cross-points on the timeline represent the first year in which the literature review portfolio introduces a keyword. Among the clusters with the earliest origins is #5 “Digitalization of urban mobility”, highlighting how technological advancements and digitalization have driven the development of MaaS services. This is evidenced by the dense connections of its keywords with other clusters, particularly #3 “Shared Mobility” and #4 “Autonomous Vehicle”. Over time, themes related to sustainability and user perspective have gained increasing attention, while the most recent developments focus on technological progress in #8 “Neural Network Approach”, #9 “New Technologies”, and #10 “Single-Leader Multi-Follower Game”.

Table 5. Categories of MaaS keyword clusters.

Thematic Group	Cluster ID	Label (LSI)	Size	Average Year	Silhouette
G1: Impact of ICT solutions on urban mobility	#5	Urban mobility	23	2019	0.775
	#8	Neural network approach	10	2022	0.936
	#9	New technologies	2	2023	0.996
	#10	Single-leader multi-follower game	3	2023	0.997
G2: Choice of transportation mode	#3	Shared mobility	28	2019	0.764
	#4	Autonomous vehicle	32	2019	0.761
	#6	Literature review *	17	2016	0.859
	#7	Risky choice	12	2021	0.867
G3: Innovative solutions for enhancing sustainability	#0	Latent class cluster analysis	31	2020	0.718
	#1	User perspective	27	2019	0.724
	#2	Sustainable mobility	25	2019	0.775

* On users’ mode-choice behavior.

provides a detailed classification of the results obtained, where the cluster cardinality represents the number of discrete keywords contained, and the silhouette score indicates how well the keywords align with the label of their respective cluster. Specifically, a silhouette score within the range of 0.7 to 0.9 or higher ensures consistency between the keywords and the clusters to which they are assigned [56].

To facilitate thematic discussions on research priorities, the 11 identified clusters have been manually classified, grouping them into three distinct categories: “Impact of information and communication technologies—ICT—Solutions on Urban Mobility” (Group 1—G1), “Choice of Transportation Mode” (Group 2—G2), and “Evaluation and Development of Sustainable Mobility Solutions” (Group 3—G3). It is important to note that, within thematic G2, since cluster #6 “Literature Review” covers a broad thematic spectrum. The focus will be specifically on studies related to modal choice behavior.

4. Thematic Discussion

The following discussion will focus on topics related to the three groups previously described (G1, G2, and G3), with the aim of providing a thorough analysis that highlights the gaps and challenges in the studies conducted so far within these areas. Special emphasis will be placed on the impact of these studies on the Sustainable Development Goals (SDGs; [3]).

4.1. Impact of ICT Solutions on Urban Mobility

Due to the need to shift mobility models towards sustainable transportation systems, technological solutions, particularly mobility apps, have gained increasing popularity. These technologies enable the aggregation of mobility options and travel information, optimize routes, and facilitate door-to-door movements. Therefore, it is essential to analyze the positive and negative implications associated with their implementation within the MaaS.

Regarding ICT, intelligent transport systems (ITS), and the internet of things (IoT), these constitute the technical backbone of smart cities [57]. As highlighted in Costantini et al. [58], within the MaaS framework, IoT creates a complex network composed of two interlinked components: on the one hand, interconnected “objects”, such as data, devices, applications, infrastructures, and various types of vehicles; on the other hand, interconnected “individuals”, represented by travelers moving across the territory to meet their mobility needs. Specifically, IoT, once fully operational and capable of maximizing its potential, will serve as the “glue” connecting the various components of the MaaS system based on artificial intelligence (AI), with a particular focus on connected and autonomous vehicles (CAV). These vehicles have the potential to blur the boundaries between public and private transportation services, making MaaS appear as a viable compromise [59]. In particular, various AI-based models have been implemented in different MaaS solutions, as in classifying drivers’ driving styles [60] and to develop optimization models to manage bike-sharing fleets in smart cities [61,62,63]. Furthermore, unlike previous studies where the impact of such solutions was not directly perceived by users, Rajabi et al. [64] introduced an innovative AI approach that ensures that MaaS service users receive a personalized itinerary, capable of adapting to individual modal preferences [65,66] and weather conditions. While the benefits of applying ICT solutions to MaaS services are evident, certain aspects require further attention. Notably, user privacy concerns related to profiling and the sharing of personal travel information—such as vehicle availability, origins, destinations, and financial information—have been identified as potential barriers to achieving MaaS objectives [67]. Within the context of MaaS, significant concerns have arisen regarding the potential for users not only to be profiled but also to be identified. For example, an individual’s movements to and from medical facilities could be correlated with specific health conditions, and destinations could include places of worship, union headquarters, political party offices, or civic organizations [58]. Despite the GDPR providing legal and technological safeguards, such as consent requirements and privacy by design, these measures alone may not be sufficient to fully protect users [68]. Beyond regulatory guarantees, there is a pressing need for increased education and awareness about the value and risks associated with personal data. Trust has been shown to be a crucial factor in driving user adoption of MaaS solutions [69]; therefore, public authorities should promote information campaigns on the risks, benefits, and opportunities arising from personal data processing for mobility services. In other words, public-private cooperation could significantly reduce the risk of unlawful data processing while simultaneously fostering the full development of MaaS services.

Furthermore, ICT solutions can contribute to digital inequality and disadvantage situations within mobility services, as not everyone is willing or able to use them [70]. Given the central role of ICT in the MaaS ecosystem and in alignment with SDG 10 of Agenda 2030, which calls for reducing inequality and ensuring social, economic, and political inclusion for all, it is necessary to assess the impact of these solutions on vulnerable social groups (VSGs), including the elderly, individuals with physical and/or cognitive disabilities, and the low-income population [71]. These categories are particularly susceptible to social exclusion. Figure 9 illustrates the complex interplay between digitalization, mobility, and social exclusion, emphasizing how digitalization, while integral to modern transportation services, can inadvertently contribute to digital inequality, particularly among VSGs. This risk is especially pronounced when mobility services are not designed with inclusive planning strategies.

Digital technologies, such as smartphones and tablets, have the potential to provide more equitable access to transportation services for people with disabilities, especially those with visual limitations [72]. ICT can also support elderly and disabled people by offering guided navigation tools that facilitate their mobility. However, many VSGs face challenges in adopting and using smartphones due to a combination of factors, including the digital divide, financial limitations, visual impairments, and a general lack of interest or knowledge in the use of advanced technological devices [73]. Additionally, people with disabilities often encounter difficulties in booking accessible vehicles, which is exacerbated by the lack of clear and available information, leading to travel restrictions [74]. Despite the introduction of the “MaaS Inclusion Index (MaaSINI)” to assess VSG inclusion within the MaaS ecosystem [75], no studies have yet presented effective solutions that feature a travel planner specifically designed to accommodate physical/cognitive limitations related to age. Such a system should generate accessible routes, provide real-time information on infrastructure status (e.g., ramps, tactile paths, elevators, stairlifts), and adjust travel times accordingly, particularly regarding ingress and egress times.

To address these challenges, future MaaS solutions must ensure universal digital access by implementing intuitive and simplified interfaces. In addition, nondigital payment methods should be integrated alongside electronic payment options to guarantee that MaaS services are accessible to all users, including those who do not have access to electronic payment methods or who choose not to use them due to privacy concerns. By adopting such inclusive strategies, MaaS can serve as a tool to actively reduce economic and social inequalities, in alignment with SDG 10, rather than contributing to their increase.

4.2. Choice of Transportation Mode

Understanding user modal preferences and their impact within the MaaS ecosystem is essential to developing effective strategies that promote the adoption of MaaS, enabling the creation of tailored and attractive mobility packages. The widespread adoption of MaaS would contribute not only to the achievement of SDG 13 “Climate Action”, but also to SDG 3 “Good Health and Well-being” with target 6, which aims to reduce the number of global deaths and injuries from road traffic accidents, through a shift towards public transport. In fact, approximately 1.19 million people die each year around the world because of road traffic crashes, which are the leading cause of death among children and young adults aged 5–29 years [76].

It is important to emphasize that traditional public transport modes, such as buses, trams, and subways, should remain the backbone of MaaS in urban areas due to their sustainability and capacity advantages [77]. However, integrating these traditional modes with faster, more affordable, and continuous services can reinvigorate public transport and offer a key solution for promoting accessibility and sustainable modes of transport while discouraging private car use [78]. Recently, Kriswardhana et al. [79] analyzed the distribution of transport modes included in MaaS packages in 14 studies conducted between 2018 and 2022, revealing that public transport and car sharing were present in all the studies reviewed. The consistent inclusion of public transport is understandable, given the evidence that users prefer mobility plans that include public transport options [80] and that public transport users are more likely to adopt MaaS services [81]. Moreover, MaaS allows the integration in the multimodal network of challenging modes, such as electric micromobility systems such as e-bikes and e-scooters, which represent sustainable mobility options, especially for specific classes of traveled distances [63,82,83,84]. Moreover, the coverage and accessibility of transit services can be completed through the implementation and promotion of these systems. Therefore, transport engineering is facing the development of new tools for supporting the forecast of potential demand both for door-to-door trips and its integration with transit.

What is surprising, however, is the inclusion of car sharing in all mobility packages, given the mixed evidence on its role and impact. While some studies have demonstrated its positive effect on increasing users’ willingness to pay [85], others have shown that instead of encouraging car owners to switch to shared services, car sharing provides access to vehicles for those who do not own one [86], thereby increasing the number of car trips. It can be highlighted that the presence of cars impacts not only the presence of vehicle flows but also road occupancy in parking. On the whole, car sharing can reduce the number of cars on the road, reducing idle time for vehicles. Even case studies present contrasting findings. For example, a pilot project in Belgium [87], which involved 73 car owners using a MaaS application for two and a half months, found that despite incentives to minimize car use (with a EUR 0.50/km penalty deducted from their MaaS package), participants displayed unsustainable behaviors. One-third of the MaaS budget was spent on personal car use, while another third went towards car-sharing services. Kamargianni et al. [88] also conducted a study with 1570 Londoners, revealing that both car owners and non-owners generally view car sharing favorably. Most participants, both car owners (51%) and non-owners (63%), agreed that car sharing is an excellent way to access a vehicle without owning one. Taking into account these findings, a balanced approach can be found in the perspective that moves away from condemning car use within MaaS and recognizes that cars are unlikely to disappear entirely. The challenge, therefore, is to find more sustainable ways to use them. “Electric Car Sharing as a Service” (ECSaaS) could be a promising solution, focusing on improving the efficiency of electric vehicle fleets [89], particularly given that private cars typically remain parked for over 90% of their lifespan. Up to now, no studies have explored the potential use of data from car-sharing trips, which could instead be utilized to enhance public transport planning, improving its competitiveness and better meeting users’ needs.

Another effective solution to improve access to urban transport services, increase vehicle occupancy rates, and reduce travel distances and emissions is demand-responsive transport (DRT). Several MaaS studies have focused on the impact of DRT in rural contexts [33,37], where organizing mobility services is particularly challenging due to long distances, low passenger volumes, and limited modal choices [90]. Some studies have proposed extending public transport from urban centers to outer, lower-density areas; however, the economic viability of such initiatives remains a major obstacle [67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91]. Eckhardt et al. [92] demonstrated the positive impact of DRT in rural areas through two pilot projects in Finland, underscoring the need for public-private partnerships (PPP) for success. These services, primarily designed for elderly people and those with reduced mobility, involved drivers assisting passengers as needed and a call center to improve inclusion for those unable or unfamiliar with mobility apps. Despite localized efforts, the integration of DRT within MaaS packages has not yet reached a satisfactory level. This gap has particularly negative implications for VSGs and rural residents, for whom DRT often represents the only viable alternative to private cars, providing essential connections to urban public transport systems.

Future research should prioritize DRT into MaaS solutions while also focusing on creating mobility packages that encourage sustainable practices, optimize the use of public transport, and accommodate car sharing in a balanced way. By ensuring accessibility for all users, MaaS can drive both environmental sustainability and road safety improvements.

4.3. Innovative Solutions for Enhancing Sustainability

A crucial consideration for mobility as a service (MaaS) concerns the effective design and implementation of incentives aimed at providing direct benefits—both economic and non-economic—to users while achieving broader sustainability outcomes for society by encouraging changes in travel behavior. To date, MaaS has typically relied on financial incentives derived from trials and venture capital, but has struggled to achieve widespread adoption [93]. To unlock the full potential of MaaS in promoting more sustainable transport options, additional strategies may be necessary.

Pricing schemes that include fare discounts have been shown to positively influence temporal commitment to mobility packages [94]. In addition, mobility packages can offer a rollover option, allowing unused credits to be transferred to the next billing cycle [80]. From a policy perspective, the European Commission has consistently promoted initiatives over the past decade to support the development of multimodal travel planning across the EU, which has contributed significantly to the rapid advancement of MaaS. In recent years, these initiatives have been bolstered by government policies aimed at promoting public transport and integrating it with private-sector technologies, leveraging advancements in ICT and IoT [34]. In particular, the political support and legislative framework provided by the Finnish government have been key factors in the success of Whim, one of the leading MaaS pilot projects in the world [95].

Furthermore, it is important to note the lack of substantial efforts to integrate private vehicles into the MaaS ecosystem, despite their dominant role in the mobility landscape. Hensher [96] suggests that shifting from a multimodal to a multiservice perspective can reverse this trend by demonstrating that the convenience of private cars must be incorporated into MaaS solutions. This can be achieved through the participation of private companies from other sectors, such as insurance companies, which can focus on customer needs without the constraints faced by traditional transport providers. This evolution represents the second generation of MaaS, known as “Mobility as a Feature” (MaaF). Insurance companies, with their extensive networks of partnerships, often provide discounts and promotions across a range of services and operate in a highly competitive market where they constantly seek to expand their market share. An optimal MaaF model might involve insurance companies acting as intermediaries, offering incentives to vehicle owners who insure with them, such as discounts on public transport or access to electric micromobility services funded by insurance premiums. Additional benefits, such as reduced annual premiums or maintenance costs, could be provided for hybrid or electric vehicles or for low annual mileage, as these factors correlate with lower accident rates [97].

The implementation and evaluation of the MaaF effectiveness will require sufficient time. However, the initial indicators are promising. Building on the lessons learned and the challenges encountered in the context of mobility as a service, the literature suggests that sustainable outcomes, aligned with the Sustainable Development Goals of Agenda 2030, can be achieved. Only through a joint effort and a comprehensive understanding of this innovative approach will it be possible to steer the future of urban mobility towards sustainability, in alignment with SDG 11 “Sustainable cities and communities”.

5. Conclusions

Many researchers and stakeholders in the transportation sector see MaaS as the mobility solution of the future, due to its potential to promote more sustainable mobility and provide a concrete alternative to private car ownership. This study conducted a systematic review of the literature to assess the current state of MaaS implementation, trends, and research gaps, with an in-depth analysis of emerging themes. Based on 819 documents selected from the WoS and Scopus databases up to 2024, a quantitative analysis of the temporal distribution of the studies was developed through bibliometric analysis, generating three types of networks that illustrate geographical analysis, collaboration relationships, and clusters identified through keyword analysis. The main findings of the bibliometric analysis are listed below:

• since 2018, the number of published papers has grown significantly, reaching a peak in 2023, with 168 publications; projections suggest a continued rise in the importance of MaaS in academic research;
• Germany, the United Kingdom, and Australia are leaders in research and international collaboration on MaaS, with Germany having the highest number of publications (104 articles), while David Hensher from the University of Sydney is the most prolific author (18 articles); the high centrality values highlight their role as key reference points in the MaaS field;
• keywords such as “numerical model”, “willingness to pay”, “transport systems”, “public transport”, and “electric vehicles” exhibited a high burst strength, aligning with the primary research themes identified through cluster analysis; three emerging thematic groups were identified for discussion, i.e., “Impact of ICT solutions on urban mobility—G1”, “Choice of transportation mode—G2”, and “Innovative solutions for enhancing sustainability—G3”.

The first part of the thematic discussion highlights those digital technologies, such as the IoT and ITS, that constitute the digital backbone of the MaaS system. However, issues concerning privacy and digital exclusion, particularly affecting vulnerable social groups, have emerged. The emphasis is on the necessity of ensuring universal access to MaaS technologies in order not to marginalize the most vulnerable populations, thus promoting the reduction of inequalities in alignment with SDG 10.

The second part examines the modal preferences of the users within MaaS while also discussing the role of each mode of transport within the ecosystem and the mobility packages. The fundamental role of public transport is reaffirmed, while the mixed evidence surrounding car sharing is discussed, with the emerging concept of “Electric Car Sharing as a Service” (ECSaaS) proposed as a point of convergence. Particular attention is also paid to demand-responsive transport, highlighting the positive impacts observed in rural projects, which have been proven to ensure inclusion for VSGs. As a result, greater promotion and integration of these services within mobility packages are recommended. This approach will also enable climate action (SDG 13) and contribute to reducing the number of global deaths and injuries from road traffic accidents (SDG 3.6).

In the final part, acknowledging the need to integrate private cars into sustainable transport systems and the failure of policies adopted so far, alignment with the second generation of MaaS, known as “Mobility as a Feature” (MaaF), is proposed. Studies conducted in this area suggest that this innovative perspective could represent a promising solution to achieve sustainable outcomes in line with Agenda 2030, fostering the creation of sustainable cities and communities according to SDG 11.

In practical terms, this study provides valuable insights and recommendations for researchers and stakeholders in the transport sector who aim to address the challenges associated with the development and implementation of MaaF, starting from the lessons learned from MaaS.

Finally, with the development of MaaS, it is not possible to ignore the risks and vulnerabilities for security on user data and system integrity, as MaaS systems use digital technologies like the Internet of Things (IoT; [98]) and intelligent transportation systems (ITS) more and more. To allow users to embrace MaaS solutions, they must be confident that personal information data is securely protected; therefore, effective cybersecurity systems are crucial. In addition to protecting sensitive data, addressing cybersecurity issues is essential in guaranteeing the general reliability and success of MaaS projects. This is an issue that has to be prioritized in research in order to provide a more thorough foundation for safe and sustainable urban mobility solutions [99]. To boost adoption and provide practical methods for gathering and using data while preserving data security and privacy, templates and instructions for sensor deployment tasks would be beneficial.

Operators’ partnerships might spearhead or assist this kind of movement, as already proposed in other significant solutions in urban and metropolitan areas [98,100]. Local government engagement is necessary to achieve transferability and comparability. The application of sensor data in planning, designing, and operating for MaaS may become more widespread if this can be accomplished in a way that unites different metropolitan areas and regions.