Chapter 8: Mobility-as-a-Service (MaaS) and New Business Models

Mobility-as-a-Service (MaaS) represents a fundamental shift in how individuals and organizations plan, book, and pay for transportation. Rather than owning or leasing private vehicles, users gain on-demand access to a curated suite of mobility options public transit, ride-hailing, bike- and scooter-sharing, car-rental, taxi services, and even micro transit through a single digital platform this new mobility ecosystem. 

.The overarching goal of MaaS is to present travel not as discrete journeys requiring multiple apps and payment methods, but as a seamless, end-to-end mobility experience optimized for cost, time, convenience, and environmental impact. This chapter introduces the key architectural components of MaaS platforms, the supporting policy and regulatory environment, and the emerging business models that underpin  

At its core, a MaaS platform relies on data integration and orchestration. Real-time feeds from public transport operators provide schedule adherence and occupancy information; shared-mobility fleets report vehicle locations, availability, and pricing; and ride-hail services expose dynamic fare estimates and wait times. Sophisticated back-end engines aggregate these heterogeneous data streams, apply routing algorithms that can blend multiple modes within a single itinerary, and deliver personalized suggestions based on user preferences whether minimizing travel time, reducing carbon emissions, or optimizing for cost. Crucially, unified payment systems handle complex fare structures, transfers, and cross-provider transactions, issuing a single monthly invoice or subscription package that covers the full spectrum of mobility services. 

Subscription-based pricing is one of the most compelling new business models enabled by MaaS. Instead of paying per trip, users select from tiered plans ranging from “all-you-can-ride” options for regular commuters to pay-as-you-go bundles that combine a set number of transit rides, shared-bike minutes, and ride-hail credits. This “mobility bundle” approach mirrors communications-as-a-service (e.g., mobile phone plans), providing cost certainty and encouraging users to make more sustainable travel choices. Operators benefit from greater demand predictability and the ability to smooth peak-period loads, while governments can incentivize off-peak usage by offering discounted “shoulder-hour” packages. 

Beyond subscriptions, dynamic pricing and demand-responsive revenue sharing are also reshaping the economics of urban transportation. In a MaaS ecosystem, service providers negotiate revenue-share agreements with the MaaS integrator. Fares can be dynamically adjusted based on real-time network conditions, raising prices during extreme congestion or reallocating discounts to underused modes and corridors. Such pricing flexibility helps balance supply and demand, discouraging car use when transit capacity is ample, or vice versa. Data-driven performance metrics on-time performance, customer satisfaction, vehicle utilization determines payouts, aligning incentives across public agencies, private fleets, and MaaS operators. 

 Platform architectures for on-demand mobility services 

Platform architectures for on-demand mobility services are designed to connect travellers, service providers, and infrastructure in a seamless, scalable, and resilient ecosystem. At the heart of these architectures lies a multi-tiered design that separates concerns across user interfaces, core service logic, data management, and integration layers. A modern on-demand mobility platform typically comprises (1) front-end applications for riders and drivers, (2) an orchestration layer that handles matching, dispatch, and routing logic, (3) a data platform that collects and processes real-time telemetry, and (4) integration adapters that connect to payment gateways, mapping services, and regulatory reporting systems. This modular decomposition enables independent scaling, iterative development, and clear separation of responsibilities. 

The front-end tier presents user experiences tailored to distinct roles. Rider apps provide intuitive booking workflows, progressive disclosure of pricing and arrival times, and real-time tracking visualizations. Driver apps focus on queue management, navigation assistance, and safety alerts. Both interfaces rely on push notifications and WebSocket connections to receive live updates from the orchestration layer. To support high availability and low latency, front ends are often deployed as single-page applications (SPAs) backed by content delivery networks (CDNs) and edge caching strategies, ensuring that core functionalities remain responsive even under heavy load or network variability.  

Beneath the user interfaces, the orchestration layer implements the business logic that powers on-demand services. Core microservices manage user authentication, profile management, and credential verification. Matching services use spatial indexing (e.g., R-trees or geohashes) to efficiently locate nearby drivers, then apply optimization algorithms such as bipartite graph matching or heuristic cost functions to assign the best match based on proximity, vehicle capacity, driver rating, and service type. Dispatch services coordinate dynamic routes by integrating with real-time traffic feeds and mapping APIs, recalculating optimal pickup and drop-off sequences as conditions change. This layer must be stateless or maintain minimal ephemeral state, allowing horizontal scaling and rapid recovery from failures. 

The data management tier underpins both operational decision-making and long-term analytics. A streaming data platforms often built on Apache Kafka or cloud equivalents captures events such as ride requests, driver status updates, and vehicle telemetry. Stream processors aggregate these events into real-time dashboards for operations centres, and feed into alerting systems that detect anomalies (e.g., sudden surge in cancellations or geographic hotspots of high demand). Simultaneously, batch ingestion pipelines land data in a data lake, where ETL jobs transform raw logs into structured datasets for demand forecasting, dynamic pricing model training, and performance reporting. Feature stores manage model inputs, ensuring consistency between training and real-time inference.