A decade of Mobility-as-a-Service research: A systematic review of modeling methods and future research agenda

Abstract

Over the last decade (2015–2025), Mobility-as-a-Service (MaaS) has rapidly evolved from a visionary concept into a mature, user-centric socio-technical ecosystem. This paper marks a ten-year methodological milestone by conducting a PRISMA-guided systematic review of 92 peer-reviewed journal articles from Web of Science, Scopus, and Google Scholar. The existing body of quantitative modeling literature informing MaaS design, operations, and regulation remains fragmented across disciplines, assumptions, and decisionmaking layers. In response, we propose a unified framework that categorizes the literature into six methodological families: simulation models, optimization models, discrete choice models, other statistical methods, data-driven and predictive machine learning models, and game theory and mechanism design models. Using this framework, we map these modeling methods onto four core MaaS research themes: demand-side modeling, supply-side operations, MaaS ecosystem governance, and platform and subscription bundle design. Major findings indicate that existing demand studies have predominantly relied on stated-preference valuations of MaaS subscription plans and bundles, with only limited revealedpreference validation; optimization models have increasingly formalized allocation, matching, and assignment under operational constraints, albeit often assuming overly simplified traveler behavior; and machine learning techniques have expanded rapidly but are generally deployed as stand-alone prediction tools rather than integrated into policy-constrained decision support systems. In addition, the maturity levels of each methodological family reveal significant disparities: foundational areas such as revealed-preference modeling and choice-based optimization are well-established (extensively studied), while emerging fields like machine learning and game theory remain less studied or in early-stage exploration. To advance the field, we provide a forward-looking agenda of 20 research directions, prioritizing more data-driven behavioral modeling, tighter demand–supply integration in operational settings, new multi-sector partnerships, and the concept of Mobility-as-a-Feature. We emphasize planning for equity and long-term impacts and the responsible incorporation of emerging technologies into next-generation MaaS. This systematic methodological review provides evidence-based guidance and a structured roadmap for researchers, operators, and policymakers, addressing identified gaps and highlighting areas requiring further development to support robust, policyaligned decision-making in MaaS.