Research
Understanding mobility through behavior, surveys, and networks.
My work develops empirical and modeling approaches for transportation research, from the quality of reported trips to mobility patterns across people, places, and modes.
Travel survey methods
Travel surveys are an important data source for analyzing travel behavior and validating transportation planning models, yet their quality is often impacted by substantial travel misreporting bias. Although passively collected big data, such as GPS trajectories, smart card records, and mobile phone signaling data, provide higher accuracy, they often suffer from sample representativeness problems and lack important background information, such as driver's license status, car ownership, activity characteristics, and household attributes. Existing methods to correct underreporting bias can be broadly classified into three categories: those based on external data (e.g., GPS trajectories and time-use surveys), those using response quality indicators (e.g., speeding, straightlining, item nonresponse, and proxy responses), and those employing behavioral models to identify and correct misreported samples. My research primarily contributes to the first two approaches.
Methods and data: household travel surveys; response-diligence indicators; behavioral models.
Related work:Zhang, Z. & Maruyama, T. (2026). Is item nonresponse associated with trip underreporting in household travel surveys? Evidence from Kumamoto, Japan. Travel Behaviour and Society, 45, 101297. DOI: 10.1016/j.tbs.2026.101297.Zhang, Z. & Maruyama, T. (2026). Immobility or soft refusal? An empirical analysis of the association between respondents’ diligence and reported immobility in household travel surveys. Transportation Research Part A: Policy and Practice, 204, 104815. DOI: 10.1016/j.tra.2025.104815.Maruyama, T., Abe, Y., Arao, S., & Zhang, Z. (2026). Increased stay-at-home or increased soft refusal? A comparison of time-use and household travel surveys in Japan. Transportation Research Procedia, 97, 151–161. DOI: 10.1016/j.trpro.2026.04.152.
Older adults' mobility
In the era of population aging, older adults' travel behavior has attracted increasing attention. Our research focuses on the heterogeneity of older adults' travel and activity patterns, reflecting the heterogeneity within the aging population. Furthermore, private cars are an important travel mode for older adults; however, many older adults have to stop driving as they get older. Understanding their travel and activity behavior after driving cessation is therefore essential for developing transportation policies and improving the well-being of older adults.
Methods and data: latent class analysis; causal analysis; travel survey data.
Related work:Zhang, Z. & Maruyama, T. (2024). Exploring the Heterogeneity of the Travel Behavior of the Elderly in the Kumamoto Metropolitan Area Based on Latent Class Analysis. In CICTP 2024, pp. 955–964. DOI: 10.1061/9780784485484.091.
Time-varying multimodal networks
Networks provide an important paradigm for representing complex systems. Like many real-world systems, transportation networks exhibit small-world and scale-free properties. Multimodal mobility systems can be naturally modeled as multilayer networks, while their strong temporal dependence makes dynamic network representations essential. Some interesting research questions include how to model heterogeneous interactions between network layers and how to identify critical components within such complex network structures. My research focuses on addressing these challenges and exploring the interpretation of the results in the transportation field.
Methods and data: complex network analysis; multimodal mobility networks; PageRank; temporal coupling.
Related work:Zhang, Z., Ando, H., Wang, Y., Zhu, T., & Yang, X. (2026). Analysis of mobility discrepancies within urban agglomerations using an extended PageRank algorithm in time-varying multimodal networks. Physica A: Statistical Mechanics and its Applications, 681, 131060. DOI: 10.1016/j.physa.2025.131060.
Built environment and urban function
The built environment significantly influences travel and activity behavior, but how is it formed? Key factors may include land-use policies and the effects of the natural environment on urban expansion and functional zoning. To be honest, this is not my area of expertise, but exploring this question with interesting collaborators has been highly rewarding.
Methods and data: QGIS and ArcGIS; statistical modeling; spatial and built-environment measures.
Related work:Manuscript in submission on ecological zones and functional layout.