**6. Conclusions**

This paper proposed a new approach to take into account the smart distribution of the transfer location impact on travel mode choice. A transformation method was proposed for mapping of the transfer locations on a two-dimensional homogeneous geocoordinate. Transformed transfer locations were grouped into six classes by the level of point density. The novel transfer location variable was

included in a mode choice model to demonstrate its underlying e ffect. The new variable was found as one of the driving determinants of mode choice. The study revealed that the transfer services in the "preferred locations" are likely to increase the smartness of the transit journey including the transfer. The transfer location and the in-vehicle travel time variables were found to be significant and uncorrelated to each other. This implies that travel direction towards transfer points may be an important factor pertaining to mode choice in addition to the travel time factor. The conventional approach of using door-to-door travel time may be not su fficient to capture the real cost of transfer.

This study provides a new approach to analyse the spatial distribution of transfer locations in relation to trip origin and destination points. This new geocoordinate technique may be useful for many applications in smart transport research including transit accessibility and connectivity studies. The conventional methods define the transit service accessibility and connectivity using a travel time constraint, where accessible areas by transit are simply defined as the travel boundary within a specific travel time period (e.g., 45 min). In a radial transit network structure, travelling to neighbouring suburbs often require a transfer at the opposite direction from the destination if there is no direct transit route connecting two suburbs, which is not so smart. For choice users (a private vehicle is available), such locations may be deemed as inaccessible by using transit. Integrating the transfer location variable to the traditional accessibility and connectivity measures may provide a more realistic representation of the service coverage of transit systems.

The findings of this study may contribute to improving the smartness of the public transit and the prediction capability of the mode choice analysis for the future transport demand. The findings presented in this study should be viewed as an exploratory e ffort to developing a new approach to account for the smartness of transfer and to test is e ffect on mode choice. The main findings will assist the transit service and performance assessment to identify service gaps and underserved areas. Identifying convenient and strategic transfer locations is essential so that scarce resources can be channelled e ffectively to improve the quality and smartness of transit service. Minimising the perceived transfer penalty will assist in increasing the competitiveness of public transport, and eventually the transit ridership. In this study, the emphasis is only given to bus journeys with a single transfer. Future research could build upon this concept to consider multimodal transit journeys and those journeys with more than a single transfer.

**Author Contributions:** Formal analysis, J.C.; Methodology, J.(B.)L. and H.H.; Writing—review and editing, J.(B.)L. and H.H. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Conflicts of Interest:** The authors declare no conflict of interest.
