**5. Conclusions**

#### **5.1 Remarks on MaaS**

Many researchers and stakeholders of the transport sector see Mobility as a Service (MaaS) as the mobility of the future. However, a lot of uncertainty lye under this travel solution. The same first statement is actually uncertain, considering that it depends on MaaS diffusion, which in turn depends on the adopted business model, on its financial convenience and on the membership rate, which in turn depend on what kind of services are offered, their level of service and their price. On the other hand, first MaaS applications have not helped to clarify the financial convenience of a MaaS.

Furthermore, MaaS is also generally associated with many virtuos impacts which can be sinthesized by saying that it goes in the direction of a sustainable mobility, by aiding and supporting intermodality. However, also this statement is not clearly confirmed by the literature. In fact, put different services in the

same market place does not suffice to guarantee intermodality: who chooses the service/services to offer to the users? Instead, the literature agrees that if MaaS is developed according to a purely commercial approach, there is a serious risk to dis-incentivize sustainable trips, encouraging instead a shift towards car use. Hence, the literature also agrees on the need of regional authorities to lead MaaS development through new models of governance. However, the way this result can be achieved is not clear.

Consistently with these premises, one of the main necessary research perspective is to investigate and identify possible regulatory frameworks, MaaS schemes and approaches for bidding/tendering the services enabling the public entity to rule the MaaS.

In addition, the way MaaS should work is clear in the literature, but models and methodologies enabling such working are very complex and largely new with respect to the approaches currently used in transportation system modelling.

From the demand side, new approaches are needed to profile users coupling with MaaS requirements and involving also psychological and social science. Moreover, new approaches are needed for the modelling of individual choices, considering the intrinsic dynamicity of the MaaS system and including the effect of service reliability and of the provided information.

From the supply side, new modelling framework are needed to integrate continuous and discontinuous services, to deal with diffused/distributed inter-modality and with dynamically changing conditions and unpredicted temporary disruptions of the service.

#### **5.2 Remarks on the Hypermode paradigm**

The widespread of real-time travel information combined with the presence of a variety of travel modes available in dense urban areas could lead travellers to reconsider their planned mode choices based on real-time events, such as realtime passenger information, transport disruptions, overcrowdings, and weather. However, most existing mode choice studies analyse the decision process as planned behaviour, and hence do not capture the influence of these real-time events on mode choices.

This paper aims to address this limitation by presenting an innovative approach to interpret mode choice, which captures an adaptive behaviour of travellers. The underpinning assumption is that the traveller first identifies a set of feasible modes that connect his/her origin to the destination; then he/she evaluates the real-time situation in order to select a specific mode of transport from the feasible set. This novel behaviour paradigm is referred to as hypermode in this paper, in analogous to the hyperpath concept used extensively in transit assignment. A two-level decisionmaking process is illustrated, which rests on planned and adaptive model choice behaviour, respectively.

A survey has been undertaken to test the proposed approach; it demonstrates the validity of the underlying assumption of hypermode, and serves as a proof of concept. As the next step of this research, we will explore different modelling approaches (e.g. Nested/Mixed Logit and discrete choice model with endogenous attribute threshold/cut-off) along with calibration and validation methods based on a wide variety of data.

For future research, the hypermode concept could be explored alongside dynamic assignment of the multimodal network, which provides feedback to mode choice models in the form of real-time events. For example, the dynamic re-distribution of shared bikes, as a result of multimodal traffic assignment, could affect the availability of bikes at docking stations and hence travellers' mode choices.

**19**

**Author details**

Stefano de Luca1

2 ITS Counsult, Bari, Italy

\* and Margherita Mascia<sup>2</sup>

\*Address all correspondence to: sdeluca@unisa.it

provided the original work is properly cited.

1 Department of Civil Engineering, University of Salerno, Salerno, Italy

© 2021 The Author(s). Licensee IntechOpen. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/ by/3.0), which permits unrestricted use, distribution, and reproduction in any medium,

*Adaptive Travel Mode Choice in the Era of Mobility as a Service (MaaS): Literature Review…*

Such a feedback mechanism between traffic assignment and model choice presents a research direction not previously investigated, and calls for a more integrated modelling approach driven by the presence of real-time travel information.

*DOI: http://dx.doi.org/10.5772/intechopen.98432*

*Adaptive Travel Mode Choice in the Era of Mobility as a Service (MaaS): Literature Review… DOI: http://dx.doi.org/10.5772/intechopen.98432*

Such a feedback mechanism between traffic assignment and model choice presents a research direction not previously investigated, and calls for a more integrated modelling approach driven by the presence of real-time travel information.
