**4. Real-world case study**

The hypermode concept is illustrated here using a real-world example. The area of interest is part of South Kensington in London.

As shown in **Figure 2**, a traveller starts his trip in O (origin) and wishes to reach the destination D. Before leaving the origin, the user has a set of feasible modes he would consider, namely bus, tube, bike-sharing, and walking, which are all accessible in the vicinity of the origin. These feasible modes are ranked by the user according to his/her own preferences, which are static in nature. For example, the traveller may consider cycling as unsafe, thus bike-sharing may receive a low rank or even is excluded from the feasible set. Moreover, the traveller usually has a

**13**

in the feasible set.

**Figure 2.**

stations nearby.

chooses to take a bus instead.

examples of potential adaptive behaviour.

*The study area in South Kensington, with available modes and routes shown.*

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

preferred mode within the feasible set, which is likely to be the one he/she pursues at the first attempt. If this preferred mode is not viable given real-time conditions (e.g. no shared bike is available, or the weather is unsuitable for walking), then the probability of selecting that mode decreases and the user will consider other modes

Based on **Figure 2**, we describe the following specific scenarios, which are

• The user includes walking and bus in his feasible mode set. He prefers to take the no.360 bus at the closest bus stop towards the destination. When he reaches the bus stop, he sees on the digital display board that the next bus will arrive in 10 minutes. Rather than waiting at the bus stop, he switches to walking knowing that the total travel time would be similar (notice that the waking

• The user, who eliminates the possibility of walking due to physical conditions, may have bus and tube in his feasible set with bus being the preferred option. Before leaving the origin, he checks his cell phone and finds out the estimated waiting time for the bus is 10 minutes. He then prefers to take the tube at the

• The user has walking, tube and bike-sharing in his feasible set with cycling being the preferred option. He approaches the nearest docking station and cannot find any available bike. In this case he decides to walk or take the tube, depending on which one ranks higher, instead of looking for other docking

• The user has walking and bus in his feasible set, with walking being his

preferred option. He is about to leave the origin when it starts raining. He then

route in this case differs from the one shown in **Figure 2**).

South Kensington Station instead of waiting for the bus.

*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*

**Figure 2.** *The study area in South Kensington, with available modes and routes shown.*

preferred mode within the feasible set, which is likely to be the one he/she pursues at the first attempt. If this preferred mode is not viable given real-time conditions (e.g. no shared bike is available, or the weather is unsuitable for walking), then the probability of selecting that mode decreases and the user will consider other modes in the feasible set.

Based on **Figure 2**, we describe the following specific scenarios, which are examples of potential adaptive behaviour.


*Models and Technologies for Smart, Sustainable and Safe Transportation Systems*

(1): Feasible set Probability of each mode to belong to the *feasible mode set* depends on:

• Health and/or environmental concern

• Real-time arrival time (bus, train)

• Weather (walking, cycling)

• Vehicle availability (bike-sharing, car club)

• Disruptions and crowdedness (bus/train/tube stations)

• Socio-economic characteristics (age, gender, income, etc.)

**Level of choice Factors affecting choice probability**

• Financial cost • Average travel time • Number of transfers (2): Final mode choice Choice probability of a specific *mode* depends on:

*The two choice levels and influencing factors in the hypermode approach.*

**12**

**Figure 1.**

**Table 1.**

**4. Real-world case study**

*Flow chart representation of the hypermode concept.*

of interest is part of South Kensington in London.

The hypermode concept is illustrated here using a real-world example. The area

As shown in **Figure 2**, a traveller starts his trip in O (origin) and wishes to reach the destination D. Before leaving the origin, the user has a set of feasible modes he would consider, namely bus, tube, bike-sharing, and walking, which are all accessible in the vicinity of the origin. These feasible modes are ranked by the user according to his/her own preferences, which are static in nature. For example, the traveller may consider cycling as unsafe, thus bike-sharing may receive a low rank or even is excluded from the feasible set. Moreover, the traveller usually has a

All of these illustrative examples have one thing in common: The pre-defined feasible modes are re-interpreted and re-ranked with the influence of real-time information, which is dynamic and stochastic in nature. This highlights the key difference between the traditional mode choice model and the adaptive behaviour that we try to demonstrate.

Note that it is possible that the repetitive occurring of a negative real-time event on a day-to-day basis may lead to the exclusion of a mode from the feasible set. For example, if a user constantly finds the bike-sharing station empty, he/she may exclude bike-sharing as one of the feasible modes in his/her planned behaviour. This, however, does not contradict the mode choice behaviour that we propose here. In fact, it still falls within the scope of the proposed two-stage decision-making process, i.e. in the forming of feasible mode choice set (see Level 1 of **Figure 1**). In most cases, the feasible mode set contains more than one element, and the realization of a particular mode choice (or sequence of mode choices) must thus rely on real-time events.

To further support the relevance and likelihood of such adaptive behaviours, we conduct a qualitative survey to validate the behavioural soundness of this subject, as described in Section 4.1.
