On-demand Mobility Services Research Articles

Quantifying the external costs of autonomous on-demand ride pooling services

Mobility On Demand (MOD) services, such as ride-pooling, provide convenient and cost-effective transportation options. While previous studies focused on operational costs and service quality, we take a broader perspective by examining the external costs associated with autonomous ride-pooling services. Incorporating external costs into the design and evaluation of MOD services enables a comprehensive understanding of their impact on the entire urban population, informing effective regulations and incentives. We present an approach for calculating space-varying external costs, accounting for factors like air pollution, climate impact, noise and accidents. These costs are integrated into FleetPy, an agent-based simulation tool for ridesharing analysis and optimization. A case study in Munich uncovers the tradeoffs between external costs, internal costs, and service quality. Our findings suggest that mid-sized vehicles with a three-person capacity strike a balance between energy efficiency and transport capacity. By applying our approach, external costs can be reduced by up to 37%.

Spatial factors associated with usage of different on-demand elements within mobility hubs: a systematic literature review

ABSTRACT Mobility hubs offer a strategic opportunity to enhance multimodal transportation and reduce greenhouse gas emissions within the transport sector. Identifying the spatial factors determinants influencing the usage of mobility hub components is essential for policymakers and transport planners aiming to identify their optimal locations. This study undertakes a systematic literature review and ranking of spatial factors associated with the usage of key on-demand services within mobility hub elements, including bike sharing, scooter sharing, car sharing, ride hailing and taxi and electric vehicle charging stations. Utilising databases such as Web of Science, Google Scholar and Francis & Taylor, we evaluated 119 records, identifying 39 key factors from social and built environments that are associated with usage of the on-demand elements. Key findings highlight the significance of population density, employment density, proximity to public transport, recreational POIs and household income, along with a noted negative association with the factor slope. This research contributes to exploring how these factors align across various on-demand mobility services that can be potentially used for policymakers, and transportation planners in evaluating potential strategies for the optimal allocation and development of mobility hubs.

Shared autonomous micro-mobility for walkable cities

The integration of autonomy into micro-mobility systems has given rise to a new solution known as shared autonomous micro-mobility (SAmM)11SAmM = Shared autonomous micro-mobility.. This paper delves into the opportunities that emerge from this concept, which is driven by the resurgence of walkable, human-centric cities and the advancements in autonomous vehicles. While most autonomous vehicle research has focused on cars and taxis, SAmM proposes more lightweight and environmentally friendly alternatives that align with the vision of more human-centered, walkable cities. At the same time, the integration of autonomy into micro-mobility systems holds the potential to enhance efficiency by reducing fleet sizes and increase convenience by transforming them into on-demand mobility services. This study presents a roadmap for the successful implementation of SAmM, covering essential aspects such as vehicle design, infrastructure requirements, and operational strategies. This research sheds light on the transformative potential of SAmM and provides insights for mobility operators, urban planners, policymakers, and researchers.

On-demand mobility-as-a-Service platform assignment games with guaranteed stable outcomes

Mobility-as-a-Service (MaaS) systems are two-sided markets, with two mutually exclusive sets of agents, i.e., travelers/users and operators, forming a mobility ecosystem in which multiple operators compete or cooperate to serve customers under a governing platform provider. This study proposes a MaaS platform equilibrium model based on many-to-many assignment games incorporating both fixed-route transit services and mobility-on-demand (MOD) services. The matching problem is formulated as a convex multicommodity flow network design problem under congestion that captures the cost of accessing MOD services. The local stability conditions reflect a generalization of Wardrop's principles that include operators’ decisions. Due to the presence of congestion, the problem may result in non-stable designs, and a subsidy mechanism from the platform is proposed to guarantee local stability. A new exact solution algorithm to the matching problem is proposed based on a branch and bound framework with a Frank-Wolfe algorithm integrated with Lagrangian relaxation and subgradient optimization, which guarantees the optimality of the matching problem but not stability. A heuristic which integrates stability conditions and subsidy design is proposed, which reaches either an optimal MaaS platform equilibrium solution with global stability, or a feasible locally stable solution that may require subsidy. For the heuristic, a worst-case bound and condition for obtaining an exact solution are both identified. Two sets of reproducible numerical experiments are conducted. The first, on a toy network, verifies the model and algorithm, and illustrates the differences local and global stability. The second, on an expanded Sioux Falls network with 82 nodes and 748 links, derives generalizable insights about the model for coopetitive interdependencies between operators sharing the platform, handling congestion effects in MOD services, effects of local stability on investment impacts, and illustrating inequities that may arise under heterogeneous populations.

Vertical Equity Analysis of Parking Reservation Based on the Auction Strategy

As an on-demand mobility service, parking reservations can greatly alleviate the issue of parking challenges. There are currently three primary strategies for parking reservation: first-come-first-served, permit reservation, and auction. In contrast to the first-come-first-served and permit-reserved strategies, the auction strategy uses dynamic pricing to allocate parking supplies efficiently based on the auction, which attracts more scholars for the research. However, parking reservations based on the auction process may have an inequity issue because drivers’ age, gender, income level, and location of residence fluctuate. This inequity may limit the growth of reserved parking by influencing parking drivers’ acceptance of reserved parking. But currently, very few scholars focus on the issue of reserved parking equity, and even fewer measure this nebulous and personal issue. In consideration of this, the Lorenz curve of parking reservation and the vertical equity index of parking reservation are proposed in this paper along with the calculation method for the index, which enables the problem of reserved parking vertical equity to be visualized and made concrete. The numerical experimental method is used to analyze the vertical equity of drivers with varying income levels, utilizing the Vickrey–Clarke–Groves (VCG) auction process as an example. According to the research, loss-averse drivers are more than gain-neutral and gain-seeking drivers when the income levels of the drivers using reserved parking are the same. With the increasing number of high-income drivers involved in the parking reservation, medium to low-income drivers would lose their chances of successful reservations because of their uncompetitive bid price which leads to inequity issues when the number is less than the number of parking spaces. In contrast, the vertical equity index changes more for gain-seeking drivers while being generally steady when loss-averse and gain-neutral drivers participate. For instance, when the profit and loss coefficient is 3 and 15% of drivers with high-income levels use the parking reservation platform, the vertical equity index rises from 0.09128 to 0.45434. The reference price has a moderating influence on the vertical equity index when the number of driver participants at high-income levels remains constant. In general, within a reasonable range, the higher the reference price, the more equitable the parking reservation procedure and the lower the vertical equity index.

A multi-task deep learning framework for forecasting sparse demand of demand responsive transit

Demand Responsive Transport (DRT) is a flexible transportation service that adjusts its routes and schedules in real time based on user requests. Real-time demand forecasting has become a timely issue since the resolution of transportation planning has increased to provide flexible services to users. Although many previous studies have been conducted to design the demand forecasting model using the mobility service historical data, there are more considerations when predicting real-time demand for on-demand mobility services. Therefore, this study aims to develop a demand forecasting model for DRT services addressing two issues. First, repeated calls for service and abandoned requests are converted into the unit of demand to capture served and potential demand. Second, given the sparsity of DRT historical data, a multi-task deep learning model is proposed to predict both the probability and volume of demand simultaneously to mitigate the data sparsity problem. A zero-inflated loss function was introduced to consider the skewed demand distribution. To attain our goals, the historical dataset of the I-MOD service trial in Yeongjongdo, Inchon, Korea was used. The performance of the proposed model is compared against several baseline models, demonstrating its superiority in forecasting demand for mobility services. The results show that a multi-task structure can alleviate the data sparsity problem by excluding where demand is not generated from the forecasting problem. The proposed models can provide insights into real-time demand forecasting with sparse demand.

Оптимізація транспортно-логістичної системи задля спрощення переміщення населення сільської місцевості

The purpose of the article is to examine existing data for the selection of survey areas, data collection for the mobile network at the level of travel chains, and analysis of travel patterns. Travel demand in three rural areas with different population densities is considered, and requirements for new mobility solutions based on demand and people's travel habits are established.
 Public transport plays a decisive role in the logistics of territorial communities of Ukraine. However, inadequate provision and implementation of technological advances have left rural areas behind and heavily dependent on private cars. This article describes how to improve the availability of transport services in rural areas, where the demand for travel is often underestimated. The demand for travel in three rural areas with different population densities was analyzed, and the requirements for new mobility solutions based on demand and people's travel habits were determined.
 There is a need to access a higher level of subsidies if on-demand mobility services cannot attract an adequate level of demand so that rural and semi-rural communities have access to the same level of services and opportunities available to urban areas. Most of these services are deployed in urban settings with high travel demand and population density. Research on regions with low demand is rarely considered.
 Many on-demand services are community-based, where demand is determined based on local knowledge and specific needs. Those mobility services that are more commercially oriented use qualitative market research to determine demand for specific population segments and specific locations.
 Social isolation is exacerbated by lack of accessibility, roads and public transport, but is also linked to socio-demographic factors.
 It is suggested that consideration be given to removing barriers to commercially viable new mobility services in rural areas by: using data-driven approaches to understand typical daily travel patterns; identifying new mobility services that can be deployed for commercial purposes and those that will require ongoing financial support to provide socially necessary services.

Outsourcing service price for crowd-shipping based on on-demand mobility services

Online-to-offline (O2O) parcel delivery service providers (PSPs) are facing ever-increasing pressure to provide timely and reliable delivery services. An emerging form of collaborative crowd-shipping based on on-demand mobility (COM) services, which utilizes the excess capacity of on-demand mobility vehicles (e.g., taxis and ride-hailing vehicles) to deliver O2O parcels, is proposed to alleviate PSP’s delivery pressure while improving the profitability for the on-demand mobility service provider (OMP). This study investigates an outsourcing service price problem for a collaborative COM considering the game play between PSP and OMP. A bilevel optimization model is formulated, and a customized iterative hybrid (IH) algorithm integrating two tabu search algorithms with objective-oriented granularity and a genetic algorithm is proposed to find the solution. Numerical experiments conducted on randomly generated instances and a real-life case study demonstrate that the COM model with a well-designed outsourcing service price can benefit both participants, with PSP achieving up to 17.0 % cost savings and OMP achieving up to 15.8% profit improvement.

Evaluating prebooked on-demand mobility services using MATSim

A common task in today’s transport simulation research is the assessment of on-demand mobility services on a large scale. A framework that is widely used for such tasks is MATSim, which has been applied frequently to study highly dynamic on-demand mobility services. To date, its efficient fleet management implementation does not yet support on-demand requests that are pre-booked in advance. This type of service, however, is seeing increased interest in the context of long-distance, rural, and inclusive mobility. Therefore, the present paper explores how prebooking can be integrated into MATSim and a use case for the city of Melun in France is presented.

A One-Way Car-Sharing Based Approach for Combined Shared Mobility of Freight and Passengers

Climate change stresses the need for research and development of innovative sustainable mobility solutions that provide reliable and convenient door-to-door services for both passengers and freight. The increase in urban population and the popularity of e-commerce further highlights the need for action. In this regard, crowd-shipping is often perceived as an efficient, cost-effective, and sustainable alternative (or complement) to the management of urban freight mobility through efficient utilization of current transportation capacities. In this framework, inspired by the concept of MaaS (Mobility as a Service) in integrating various forms of transport and transport-related services into a single on-demand mobility service, this paper proposes a car-sharing-based service for the combined mobility of passengers and freight. In doing so, one-way car-sharing and crowd-shipping concepts are integrated in order to serve part of the existing freight demand in a sustainable and cost-efficient way for users, societies, and the environment. An optimization model is proposed to optimally plan the activation of one-way car-sharing and crowd-shipping services and to determine the optimal number of vehicles to assign to them. Such decisions are aimed at minimizing the total imbalance by serving passenger and freight demand during different time periods. In doing so, the willingness of users to carry freight in their vehicles is also taken into consideration. The capability of the proposed approach is evaluated through representative numerical examples aimed at showing the impact of the model parameters on the solution.

What drives the joint demand for ride-hailing and carsharing services? Understanding consumers’ behaviors, attitudes, & concerns

Driven by the emerging collaborative consumption trends, new shared ownership-based business models provide more flexible and accessible on-demand mobility options. This study simultaneously analyzes factors correlated with the consumers’ use of two interrelated disruptive on-demand mobility services, including ride-hailing (RH) and carsharing (CS). A comprehensive behavioral framework is presented to explicitly address important methodological concerns regarding the complex stochastic dependence between the use of CS and RH programs and the underlying behavioral heterogeneity in latent factors influencing the use of the two shared mobility options. Using unique data from the California Vehicle Survey, a rigorous elliptical and Archimedean copula-based finite mixture bivariate ordered probit (BOP) modeling methodology is used to understand behavioral, attitudes, and concern-related correlates of households’ participation levels in CS and RH programs. Characterized by a clayton-copula based finite mixture BOP model, participation levels in CS and RH programs exhibited complex (both synergistic and competing) non-linear stochastic dependence patterns. With a stronger left tail dependence, majority of the households having lower levels of participation in RH also had lower participation levels in CS programs. Contrarily, of the households with higher levels of participation in RH programs, the majority had lower levels of participation in CS programs (revealing weaker right tail dependence). Taste heterogeneity was observed in the unobserved determinants of CS use. Results show that current users of CS and/or RH programs tend to be those with greater awareness, owners of plug-in electric vehicles, those making greater transit trips, workers with greater commute distances, and those lacking free parking at their residences. Compared to ride-hailing, the negative effect of more frequent drivers in a household on CS use was less pronounced. Analysis of marginal and joint marginal effects provided deeper insights into the (interactive) effects of other behavioral and socio-demographic correlates. Treatment effects were simulated and discussed to further demonstrate the policy implications of our results. The new empirical insights provide a more granular understanding of the use patterns of on-demand shared mobility services and can inform more context-sensitive travel forecasts for planning and programming purposes.

Optimising fleet sizing and management of shared automated vehicle (SAV) services: A mixed-integer programming approach integrating endogenous demand, congestion effects, and accept/reject mechanism impacts

Shared automated vehicles (SAV) are expected to benefit the sustainable development of urban regions and alleviate the negative impacts brought by the increasing number of private cars. In this paper, we envision a future scenario where non-pooled SAVs replace private cars and provide public on-demand mobility services to satisfy the mobility needs of a city’s residents. To help service providers make profitable fleet sizing and management decisions, we develop a mixed-integer non-linear programming model that considers the congestion effects and the mode choice of urban travellers in different income classes, between SAVs and bicycles. Our model optimises both strategic decisions (fleet size, initial fleet distribution, and service quality level) and operational decisions (trip assignment, vehicle routing, parking, and relocation). Travellers’ preference for both transport modes is described through a binary logit model and congestion effects are described by dynamically varying travel times with respect to traffic flow in a non-linear fashion. In addition, we investigate two types of accept/reject mechanisms (mandatory vs. non-mandatory acceptance) which lead to an endogenously determined acceptance rate that can affect travellers’ willingness to use SAV services. The computational challenge posed by the non-linear and non-convex nature of the model is addressed through reformulation and the use of outer-inner approximation methods combined with a breakpoint generation algorithm. We demonstrate the effectiveness of our proposed method in a case study of the city of Delft in The Netherlands, as well as a scaling analysis on three toy networks with various sizes and demand profiles. A sensitivity analysis of key parameters is carried out to assess system performance. Computational results indicate that fleet sizing decisions are influenced not only by the population’s geographical distribution and land use patterns but also by the pricing strategy, unit operating costs of the SAV fleet, network congestion level, and traveller behaviour. When the price rate of using SAVs is low, the fleet sizing decisions can also be influenced by the trip accept/reject mechanism and the travellers’ sensitivity to the service quality level. In addition, a low price of SAV service will attract more users but may not necessarily bring a higher profit because of the increased traffic congestion.

Unraveling behavioral factors influencing the adoption of urban air mobility from the end user's perspective in Tehran – A developing country outlook

The integration of shared, autonomous, electric, and on-demand mobility services introduces urban air mobility (UAM) as an emerging intra-city transportation service operated by vertical take-off and landing aircraft. Despite the potential benefits of UAM limited literature exists on the end user's perception and the challenges to its societal acceptance, especially in developing economies. This paper addresses this gap in the literature by investigating the factors influencing UAM usage intention in Tehran, a developing country capital. Within the framework of the extended technology acceptance model, the study collected 828 valid questionnaires through a face-to-face survey. Structural equation modeling is employed to highlight the effects of a wide range of factors associated with individual differences, system characteristics, and social influence. The findings indicate that attitude, subjective norms, trust, perceived usefulness, perceived enjoyment, personal innovativeness, ecological awareness, and price evaluation are positively associated with UAM usage intention, while perceived safety correlates with the attitude toward UAM negatively. These findings provide valuable insights for policymakers and industry stakeholders, fostering a deeper understanding of end-user perspectives in developing countries, and prudently promising the potential for its targeted implementation. The study suggests that although its widespread adoption in developing countries is unlikely to occur in the near future, UAM should be regarded as a viable early and timely solution to persistent issues such as chronic congestion in metropolitan areas of the global south.

A deep real options policy for sequential service region design and timing

There is a need for strategic flexibility in well-timed deployment decisions in the design and timing of mobility service regions, i.e. cast as ‘real options’. This problem becomes increasingly challenging considering the multiple interacting real options in such deployments. We propose a scalable machine learning (ML) based real options (RO) framework for multi-period sequential service region design and timing problem for mobility-on-demand (MoD) services, framed as a Markov decision process with non-stationary stochastic variables. A value function approximation policy from the literature uses multi-option least squares Monte Carlo simulation to get a policy value for a set of interdependent deployment decisions as deferral options (i.e., CR policy). The objective is to determine the optimal selection and timing of a set of zones to include in a service region (which to add now, which to defer to a later time to reconsider). However, prior work required explicit enumeration of all possible sequences of deployments (i.e. 5 options translate to 120 sequences). To address the combinatorial complexity arising from sequence enumeration, we propose a new variant ‘deep’ real options policy using an efficient recurrent neural network (RNN) based ML method (i.e., CR−RNN policy) to sample sequences to forego the need for enumeration, making the network design and timing policy tractable for large scale implementation. Experiments based on multiple service region scenarios in New York City demonstrate the efficacy of the proposed policy in substantially reducing the overall computational cost (i.e., time reduction associated with the RO evaluation of more than 90% of total deployment sequences is achieved), with zero to near-zero gap compared to the benchmark. We validate the model in a case study of sequential service region design for expansion of MoD services in Brooklyn, NYC, under service demand uncertainty. Results show that using the CR-RNN policy in determining optimal real options deployment strategy yields a similar performance (≈ 0.5% within the CR policy value) with significantly reduced computation time (about 5.4 times faster).

Locating Method for Pick-up and Drop-off Spots for Shared Autonomous Vehicle-based Mobility Services

To provide safety to users and not disturb traffic flow, autonomous vehicles in shared on-demand mobility services cannot stop everywhere; thus, pick-up and drop-off (PUDO) spots must be dedicated or established for them. Our research objective is to propose a method to locate the PUDO spots for shared autonomous vehicle-based mobility services as this topic has been barely studied. The number of required PUDO spots is calculated, and the location is chosen considering the urban environment, walking radius, vehicle occupancy, and time for boarding and alighting among other parameters. Different from some methods applied to shared mobility, we consider the existing infrastructure (e.g., parking spaces) as potential locations for PUDO spots. The method is applied to a study case, demonstrating the applicability and providing the main findings: (i) the required number of PUDO spots decreases if willingness to walk increases; (ii) with a 3-min walking radius, 83% of curbside parking spaces can be repurposed, and 100% is reached with a 10-min walking radius; (iii) the minimum of 55% of curbside parking spaces can be repurposed with 10-min walking radius and without locating PUDO spots in private parking. Using our method, cities can determine the quantity of PUDO spots and their locations, being prepared in advance for the required changes in the existing infrastructure as well as the freed-up space to be repurposed.

The distributed ownership of on-demand mobility service

Mobility as a service (MaaS) is an innovative and value-laden method to operate on-demand mobility services within the existing transport infrastructures. A big challenge for the transport system is that the transport infrastructure cannot be restructured as quickly as a digital platform like cloud computing infrastructure services, which is a constraint that limits the balance between mobility demand and supply in real-time. The ownership of mobility services needs to be redefined to establish legal binds between service vendors and commuters, further demonstrating the capacity to balance the demand and supply of mobility services in real-time. The ownership of a mobility service is defined as the commuter's right to use the mobility service provided by the service vendor under agreed terms and conditions. This is facilitated by MobiToken in this study. An eMarket-place is proposed for commuters, mobility service vendors and other stakeholders to exchange MobiToken, where the marketplace demonstrates the potential use of blockchain tools, methods and algorithms, where a specialised capability for reliably link stakeholders is proposed through a nonrevertible blockchain dataset. The bidden price of MobiToken reflects the demand for mobility services in the real world, where congestion charges are incorporated within the bidden price. The performance of the preliminary platform is evaluated with three testing scenarios, and the results of performance testing indicate the stability and reliability of the Raft consensus algorithm for the blockchain database.

Assessment of drivers and barriers in the adoption of Mobility as a Service (MaaS): a case study of Noida, India

ABSTRACT Mobility as a Service (MaaS) has emerged as a transformative concept in urban transportation, integrating multiple modes of transportation through digital platforms to provide seamless travel experiences. By utilizing smartphone applications, MaaS simplifies trip planning, real-time information access, and consolidated payment systems, offering convenience to users. However, MaaS research has primarily focused on developed countries with well-established transport systems, making it crucial to explore its potential and challenges in developing cities of the global south, such as Noida in India. In this study, Noida, a satellite town of NCT Delhi, was chosen as only case study to gather data on user behaviour and preferences regarding app-based mobility services. A comprehensive survey collected information on socio-economic factors, personal vehicle ownership, commuting patterns, public transport usage, and attitudes towards digital ecosystems and app-based mobility services. Principal Component Analysis and K-means Clustering techniques were applied to identify distinct user types and categories, providing insights into user preferences and expectations. The analysis of the collected data revealed user clusters and their respective characteristics. The sustainability aspects of on-demand mobility services were evaluated, comparing user perceptions with private vehicle usage. The study also examined the impact of app-based mobility services on public transport and identified barriers and constraints specific to different user clusters, contributing to a better understanding of the feasibility of implementing MaaS. The findings will provide valuable insights for policymakers and transportation authorities, enabling the development of strategies and interventions to enhance urban mobility and foster MaaS adoption. By addressing the specific needs and preferences of users, MaaS can play a significant role in improving the efficiency and sustainability of transportation in complex urban environment cities like Noida in India.

An economic analysis of incorporating new shared mobility into public transportation provision

Transit agencies in the US have shown great interests in the possibility of incorporating on-demand shared mobility modes into their fixed-route transit services. However, the cost-effectiveness of on-demand modes has not been clearly demonstrated, and there lacks an effective method for transit agencies to compare the costs of different service provision options. This study develops an economic-theory-based framework that appropriately conceptualizes the total economic cost of incorporating on-demand modes into transit. Based on the theoretical framework, a simulation model is built to operationalize an approach for evaluating the cost-effectiveness of transit-supplementing, on-demand mobility services. We demonstrate the applicability of this approach using Via to Transit program in the Seattle region. By accounting for both the service provider's cost and the users' cost, we obtain a more complete and accurate measure for the cost advantages of the on-demand modes in this case in comparison to expanding fixed-route transit, where the total economic cost for the on-demand mode is 22% lower than the fixed route transit. The theoretical framework and the simulation model can support the decision-making of public transit agencies as they explore incorporating mobility on demand to supplement traditional transit.

An optimization model of on-demand mobility services with spatial heterogeneity in travel demand

We propose an optimization model that predicts the required fleet size of on-demand mobility (ODM) services at various service levels for a range of demand patterns. Both customer waiting time and spatial heterogeneity in demand are considered when formulating the flow conservation equation for the system’s steady state. The demand–supply imbalance problem is solved through vehicle relocation. Vehicle relocation is reformulated into a simple linear programming problem, and the objective function is set to minimize the number of relocation vehicles. Furthermore, indicators were defined to more effectively explain the effects of spatial heterogeneity of travel demand on the required fleet size. The analysis results show that even among regions with the same average demand rate, the required fleet size may vary depending on the spatial demand pattern of origins and destinations. The higher the ratio of long-distance travel and the greater the need for vehicle relocation, the larger the fleet size required. Furthermore, the user travel time reduction effect is maximized at the minimum fleet size and then steadily decreases as fleet size increases. Cost minimization requires more vehicles than fleet size minimization. Cost minimization may be an advantageous goal for operation of an ODM service as a driverless vehicle service. The simultaneous consideration of both fleet size minimization and cost minimization provides a required fleet size range. Although the proposed model is simple and includes approximations, it helps provide insight into ODM service design and investment decision-making. Thus, the proposed model can be used by governments or agencies to design policy or operationally optimized services. Moreover, the model can serve as the basis for the practical implementation of planned services.

An integrated ride-matching and vehicle-rebalancing model for shared mobility on-demand services

Shared mobility on demand (MoD) services are receiving increased attention as many high-volume ride-hailing companies are offering shared services (e.g. UberPool, LyftLine) at an increasing rate. Also, the advent of autonomous vehicles (AVs) promises further operational opportunities to benefit from these developments as AVs enable a centrally operated and fully connected fleet. There are two fundamental tasks for a shared MoD service: ride-matching and vehicle-rebalancing. Traditionally, these two tasks are optimized sequentially and independently. The paper formulates an integrated ride-matching and vehicle-rebalancing problem for shared MoD services which simultaneously optimizes these two tasks. We propose a graph-based methodology to solve the integrated ride-matching and vehicle-rebalancing problem with a novel rebalancing cost term quantifying supply contributions of vehicle scheduling to zonal supply deficit balances (deviations from the desired supply level) in the network. The integrated model performance is validated using a large-scale empirical shared MoD dataset by comparing with state-of-the-art sequential models. Generally, the integrated model improves the level of service and sustainability performance compared to the sequential model. The detailed analysis shows that the vehicle rebalancing in the integrated model is replaced by a more effective ride-matching and penalizing singly served trips in the integrated model can further improve its sustainability performance.