International Choice Modelling Conference, International Choice Modelling Conference 2017

Font Size: 
Flexibility or uncertainty? Predicting modal shift towards demand responsive public transport
Maria J Alonso Gonzalez, Niels van Oort, Oded Cats, Serge Hoogendoorn

Last modified: 28 March 2017

Abstract


The urban transportation sector is undergoing significant changes. Firstly, the shared-economy paradigm has led to both car sharing schemes and Transportation Network Companies (TNCs, such as Uber or Lyft). And secondly, on-going technology developments are underway to make autonomous driving a reality.

However, individual, even if not privately owned, modes of transport are unlikely to be able to cope with all the transport demand in urban areas. The public transportation sector can potentially combine collective transport with the latest societal and technological innovations leading to collective demand responsive transport (DRT). Demand responsive transport can be described as a collective on-demand mode of transport which provides flexible mobility in time and possibly also in space.

In the last decades, DRT was primarily conceived as a mode of transport for rural areas or to assist the elderly and the disabled. Lately, however, research studies have started considering the feasibility of this mode of transport for urban settings. Recent investigations of DRT have located their case studies in Lisbon (Portugal) (Martinez et al., 2015) and Hino (Japan) (Atasoy et al., 2015). These studies focus mainly on the  algorithms behind such a system. Helsinki (Finland) even implemented a real pilot between 2012-2015. Nevertheless, there is still lack of knowledge on how the shift away from fixed schedules and routes is perceived by the users. Are those aspects perceived positively as injecting flexibility to the system (“I am not fixed to a timetable”) or are they perceived negatively as inducing uncertainty to the trip ( “Will I get a suitable match next time?”)?. Our study aims at providing a broader understanding of the fuzzy variable time inherent to DRT in the mode choice decision.

Different studies have addressed mode shift including DRT services and reliability aspects. Ryley et al. (2014) includes the deviation from the expected starting time of the trip as one of the attributes when comparing bus or car to DRT. Khattak et al. (2004) asks respondents to rate different reliability and flexibility statements concerning DRT. In Diana (2010), respondents rate their mode shift propensity and different cognitive and affective modal attitudes. However, none of the mentioned approaches quantifies the different flexibility-uncertainty characteristics that surround DRT and includes these perceptions in the overall SP experiment. Our approach is expected to shed light into how the time attributes that appear in DRT influence mode choice. Next to the flexible-uncertain characteristics surrounding DRT, our study aims at studying the perception of DRT versus conventional public transport as well as examining mode choice preference against the car and individual on-demand services (Taxi and Uber).

To deal with these questions, a stated preference (SP) experiment is performed. To avoid overloading respondents, the new attributes related to flexibility-uncertainty that appear in the DRT mode will be understood as part of a time construct in an HII (Hierarchical Information Integration) experiment. HII is a method for handling multi-attribute judgement problems with a large number of attributes; in it, a logical decomposition of the decision problem takes place (Louviere, 1984). The HII can be seen as a separated SP experiment in which different levels of the variables that define the construct are rated. This methodology makes it easier for abstract concepts (such as the flexibility-uncertainty construct in our research) to be quantified. The HII experiment will be quantified in a 9 point scale. The outcome of our HII experiment will then be incorporated in the higher-level mode choice experiment as a flexibility-uncertainty attribute of DRT. Richter et al. (2012) also included HII in their mode choice SP experiment to analyse quality of connection, comfort and information.

To answer our research questions, a web-based survey among the Dutch population is designed, conducted and analysed. In the Netherlands, 91% of the households have access to the internet and 81% of the Dutch population uses internet daily or almost daily (CBS, 2016). Hence, a web-based survey is considered sufficient to get a representative sample of the population. A stated preference survey, as opposed to revealed preference survey is performed since the envisioned DRT system is not yet established. Jain et al. (2017) reviews existing studies on DRT services and identifies shopping and social trips as the most recurrent trip purposes for the use of DRT systems. Based on this finding, our SP is explained in a context of a free-time trip (as opposed to a business or a commuting trip). We also fix the context to an urban/suburban context (as opposed to rural context), since we want to study the impact that DRT services can have in the city. The addressed variables are walking time, waiting time, in-vehicle time, travel cost and the constructs resulting from the HII experiment. In addition to the SP choice experiment, the survey includes socio-economic variables, trip routines and questions concerning the inclination to use DRT. These will be compared with the values of the Netherlands Mobility Panel (MPN). The MPN is a yearly survey performed nationwide that includes a household questionnaire, an individual questionnaire and RP data of a 3-day travel diary (Hoogendoorn-Lanser et al., 2015). Biogeme is used for values estimation and SPSS to analyse the surveys’ responses. A NL model within the RUM paradigm is taken as the appropriate model to be estimated since the NL model overcomes the restrictive MNL assumption of the independence of irrelevant alternatives (IIA).

This study expects to increase the understanding of the time flexibility-uncertainty that surrounds DRT. The results of this study will enable the analysis of potential modal shift and the extent of travel generation and substitution in the event that a DRT service is offered. By testing various nested structures, the correlations of DRT with conventional public transport and individual travel modes can be identified.

 

Atasoy, B., Ikeda, T., Song, X., & Ben-Akiva, M. E. (2015). The concept and impact analysis of a flexible mobility on demand system. Transportation Research Part C, pp 373-392.

CBS. (2016). ICT, kennis en economie 2016. The Hague, Netherlands: Centraal Bureau voor de Statistiek.

Diana, M. (2010). From mode choice to modal diversion: A new behavioural paradigm and an application to the study of the demand for innovative transport services. Technological Forecasting & Social Change 77, 429 - 441.

Hoogendoorn-Lanser, S., Schaap, N. T., & OldeKalter, M.-J. (2015). The Netherlands Mobility Panel: An innovative design approach for web-based longitudinal travel data collection. Transportation Research Procedia 11, 311 - 329.

Jain, S., Nicole Ronald, R. T., & Winter, S. (2017). Predicting susceptibility to use demand resopnsive transport using demographic and trip characteristics of the population. Travel Behaviour and Society, pp 44-56.

Khattak, A. J., & Yim, Y. (2004). Traveler response to innovative personalized demand-responsive transit in the San Francisco Bay Area. Journal of urban planning and development 130.1, pp 42-55.

Louviere, J. J. (1984). Hierarchical Information Integration: a New Method For the Design and Analysis of Complex Multiattribute Judgment Problems. NA - Advances in Consumer Research Volume 11, eds. Thomas C. Kinnear, Provo, UT: Association for Consumer Research, (pp. 148 - 155).

Martinez, L. M., Correira, G. H., & Viegas, J. M. (2015). An agent'based simulation model to assess the impacts of introducing a shared-taxi system: an application to Lisbon (Portugal). Journal of Advanced Transportation, pp 475-495.

Richter, C., & Keuchel, S. (2012). Modelling Mode Choice in Passenger Transport with Integrated ierarchical Information Integration. Journal of Choice Modelling, pp 1-21.

Ryley, T. J., & Peter A. Stanley, M. P. (2014). Investigating the contribution of Demand Responsive Transport to a sustainable local public transport system. Research in Transportation Economics, pp 364-372.

 

 


Conference registration is required in order to view papers.