Tomoki Nishi
Research Subjects
Tomoki Nishi has a total of 20 co-authors. Here are the top co-authors by number of shared publications:
- Takahiro Shiga
Social Systems Research-Domain, Toyota Central R&D Labs., Inc., 41-1, Yokomichi, Nagakute, 4801192, Aichi, Japan
3 shared publications - Atsuo Komatsubara2 shared publications
- Ken Hidaka
Social Systems Research-Domain, Toyota Central R&D Labs., Inc., 41-1, Yokomichi, Nagakute, 4801192, Aichi, Japan
2 shared publications - Shintaro Fukushima
Toyota Motor Corporation, Tokyo, Japan
2 shared publications Takeyuki Sasai
2 shared publications- Aoi Watanabe1 shared publication
Latest Publications
- 26 July 2026
- Published by SAGE Publications in Transportation Research Record: Journal of the Transportation Research Board
Discrete Choice Model with Generalized Additive Utility Network for Interpretable Policy Evaluation
Discrete choice models (DCMs) are widely used for prediction and policy analysis, but standard multinomial logit (MNL) models with linear utility may fail to capture nonlinear response patterns in travel behavior. Neural extensions, such as deep neural network with alternative-specific utility functions (ASU-DNN), improve predictive flexibility, but their learned utility representations can be difficult to interpret and may be less stable under counterfactual shifts. This study proposes the generalized additive utility network (GAUNet) and its interaction variant, generalized additive and interactive utility network (GAIUNet), which represent systematic utility as a low-dimensional additive decomposition of learned neural shape functions. The proposed models are evaluated using synthetic transport mode choice data and real-world probe-person data collected in Tokyo, Japan, from 2018 to 2021. In synthetic experiments, GAUNet achieves predictive performance comparable to ASU-DNN and better than linear MNL, while yielding more stable policy evaluation under counterfactual changes in travel cost and access time. In real-world data, GAUNet and GAIUNet outperform linear MNL and remain competitive with ASU-DNN, while producing utility shapes that are easier to inspect at the attribute level. The results suggest that the proposed framework can recover interpretable nonlinear utility patterns without requiring cutoff locations to be specified in advance, thereby providing a transparent and flexible alternative to more entangled neural DCMs.
- 26 July 2026
- Published by SAGE Publications in Transportation Research Record: Journal of the Transportation Research Board
Discrete Choice Model with Generalized Additive Utility Network for Interpretable Policy Evaluation
Discrete choice models (DCMs) are widely used for prediction and policy analysis, but standard multinomial logit (MNL) models with linear utility may fail to capture nonlinear response patterns in travel behavior. Neural extensions, such as deep neural network with alternative-specific utility functions (ASU-DNN), improve predictive flexibility, but their learned utility representations can be difficult to interpret and may be less stable under counterfactual shifts. This study proposes the generalized additive utility network (GAUNet) and its interaction variant, generalized additive and interactive utility network (GAIUNet), which represent systematic utility as a low-dimensional additive decomposition of learned neural shape functions. The proposed models are evaluated using synthetic transport mode choice data and real-world probe-person data collected in Tokyo, Japan, from 2018 to 2021. In synthetic experiments, GAUNet achieves predictive performance comparable to ASU-DNN and better than linear MNL, while yielding more stable policy evaluation under counterfactual changes in travel cost and access time. In real-world data, GAUNet and GAIUNet outperform linear MNL and remain competitive with ASU-DNN, while producing utility shapes that are easier to inspect at the attribute level. The results suggest that the proposed framework can recover interpretable nonlinear utility patterns without requiring cutoff locations to be specified in advance, thereby providing a transparent and flexible alternative to more entangled neural DCMs.
- Open Access
- 17 July 2026
- Published by Elsevier in SSRN Electronic Journal
Lock-in or Shift? The Direction of Urban Mode Share Change under Nonlinear Tipping-Point Patterns: Evidence from 1-km Grids in Three Japanese Metropolitan Areas
While modal shift is increasingly expected as a demand-side measure for climate change mitigation, progress has been limited in many cities. This stagnation has been linked to nonlinear, self-reinforcing dynamics in urban mode shares, including thresholds and tipping points. However, prior studies have not fully captured such dynamics due to coarse spatial resolution and the widespread use of models that assume linear responses. To address this gap, we present an analytical framework that integrates high-resolution (1-km grid) data with a nonlinear discrete choice model to characterize spatiotemporal changes in urban mode shares. Using evidence from the three major metropolitan areas in Japan, we show that the direction of decadal mode-share change (i.e., which modes increase or decrease) is highly sensitive to nonlinear effects in the utility specification. Applying the model by area clusters further reveals that tipping-point structures in car-related mode shares vary substantially with local characteristics. While the tipping points estimated for car-related modes are broadly consistent with prior findings, clusters closer to city centers show the potential to reduce car dependence even when the initial car mode share exceeds the tipping-point threshold (30%). In addition, a model-based counterfactual policy analysis that increases railway service frequency suggests that policy impacts differ markedly across clusters, with larger effects in high-density residential areas near the urban core. Overall, our results highlight the importance of explicitly accounting for intra-urban nonlinearity and tipping-point structures when identifying where policy impacts can be maximized and designing spatially targeted transport policies.
- 17 July 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessLock-in or Shift? The Direction of Urban Mode Share Change under Nonlinear Tipping-Point Patterns: Evidence from 1-km Grids in Three Japanese Metropolitan Areas
While modal shift is increasingly expected as a demand-side measure for climate change mitigation, progress has been limited in many cities. This stagnation has been linked to nonlinear, self-reinforcing dynamics in urban mode shares, including thresholds and tipping points. However, prior studies have not fully captured such dynamics due to coarse spatial resolution and the widespread use of models that assume linear responses. To address this gap, we present an analytical framework that integrates high-resolution (1-km grid) data with a nonlinear discrete choice model to characterize spatiotemporal changes in urban mode shares. Using evidence from the three major metropolitan areas in Japan, we show that the direction of decadal mode-share change (i.e., which modes increase or decrease) is highly sensitive to nonlinear effects in the utility specification. Applying the model by area clusters further reveals that tipping-point structures in car-related mode shares vary substantially with local characteristics. While the tipping points estimated for car-related modes are broadly consistent with prior findings, clusters closer to city centers show the potential to reduce car dependence even when the initial car mode share exceeds the tipping-point threshold (30%). In addition, a model-based counterfactual policy analysis that increases railway service frequency suggests that policy impacts differ markedly across clusters, with larger effects in high-density residential areas near the urban core. Overall, our results highlight the importance of explicitly accounting for intra-urban nonlinearity and tipping-point structures when identifying where policy impacts can be maximized and designing spatially targeted transport policies.
- Open Access
- 7 July 2026
- Published by Elsevier in SSRN Electronic Journal
Do regional climatic conditions influence private passenger vehicle ownership and mileage?
Regional characteristics of travel behavior have been widely investigated, but the influence of regional climatic conditions on passenger vehicle ownership and mileage has received limited attention. To address this research gap, this study statistically assesses the effects of ten-year mean meteorological conditions on the ownership rate (number of vehicles per capita) and mean mileage (kilometers per vehicle per year) of passenger vehicles using data for 1,654 municipalities in Japan. The results indicate that air temperature, precipitation, snow cover, and sunshine duration are associated with regional differences in the ownership rate and mean mileage. Higher ownership rates and mean mileage of standard cars are observed in colder municipalities, whereas a higher number of days with heavy snow cover is associated with lower ownership rates and mean mileage regardless of vehicle category. These findings suggest that regional climate is a factor contributing to regional differences in vehicle ownership and use.
- 7 July 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessDo regional climatic conditions influence private passenger vehicle ownership and mileage?
Regional characteristics of travel behavior have been widely investigated, but the influence of regional climatic conditions on passenger vehicle ownership and mileage has received limited attention. To address this research gap, this study statistically assesses the effects of ten-year mean meteorological conditions on the ownership rate (number of vehicles per capita) and mean mileage (kilometers per vehicle per year) of passenger vehicles using data for 1,654 municipalities in Japan. The results indicate that air temperature, precipitation, snow cover, and sunshine duration are associated with regional differences in the ownership rate and mean mileage. Higher ownership rates and mean mileage of standard cars are observed in colder municipalities, whereas a higher number of days with heavy snow cover is associated with lower ownership rates and mean mileage regardless of vehicle category. These findings suggest that regional climate is a factor contributing to regional differences in vehicle ownership and use.
- Open Access
- 1 April 2026
- Published by Elsevier in Journal of Cleaner Production
- Vol. 554
Impacts of critical minerals on the diffusion of light-duty battery electric vehicles under endogenous price–demand feedbacks
- 1 April 2026
- Published by Elsevier in Journal of Cleaner Production
- Vol. 554
Open AccessImpacts of critical minerals on the diffusion of light-duty battery electric vehicles under endogenous price–demand feedbacks
- Open Access
- 21 November 2025
- Published by Elsevier in Transportation Research Part D: Transport and Environment
- Vol. 150
Charging behavior differences between BEVs and PHEVs: Evidence from vehicle log data
- 21 November 2025
- Published by Elsevier in Transportation Research Part D: Transport and Environment
- Vol. 150
Open AccessCharging behavior differences between BEVs and PHEVs: Evidence from vehicle log data
- Open Access
- 15 July 2025
- Published by Elsevier in Transportation Research Part D: Transport and Environment
- Vol. 146
How to improve PHEV electric mileage ratios? Factor decomposition with Explainable AI
- 15 July 2025
- Published by Elsevier in Transportation Research Part D: Transport and Environment
- Vol. 146
Open AccessHow to improve PHEV electric mileage ratios? Factor decomposition with Explainable AI
- Open Access
Changing role of battery electric vehicle charging strategies in decarbonizing Japanese power systems
- 8 October 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 2146-2153
High Density Automated Valet Parking via Multi-agent Path Finding
Improving the land use of parking lots is an essential component of good urban planning. To this end, many planners have introduced cooperative automated valet parking (Co-AVP) systems that realize high-density parking (HDP). Effective HDP requires the relocation of vehicles that blocks other vehicles from entering and leaving the Co-AVP. This paper proposes and solves a destination-assignment and path-planning problem for multiple vehicles with three roles: entering, leaving, and relocating. We minimize the total travel time of leaving vehicles and the total travel distance of entering and relocating vehicles from a service-policy viewpoint. We then propose two methods, integer linear programming (ILP)-based and conflictbased search (CBS)-based, for solving the problem. Numerical experiments demonstrated that the ILP-based approach scales better than the CBS-based approach. We also confirmed the applicability of our approaches to realistic parking situations.
- 8 October 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 2146-2153
High Density Automated Valet Parking via Multi-agent Path Finding
Improving the land use of parking lots is an essential component of good urban planning. To this end, many planners have introduced cooperative automated valet parking (Co-AVP) systems that realize high-density parking (HDP). Effective HDP requires the relocation of vehicles that blocks other vehicles from entering and leaving the Co-AVP. This paper proposes and solves a destination-assignment and path-planning problem for multiple vehicles with three roles: entering, leaving, and relocating. We minimize the total travel time of leaving vehicles and the total travel distance of entering and relocating vehicles from a service-policy viewpoint. We then propose two methods, integer linear programming (ILP)-based and conflictbased search (CBS)-based, for solving the problem. Numerical experiments demonstrated that the ILP-based approach scales better than the CBS-based approach. We also confirmed the applicability of our approaches to realistic parking situations.
- 20 March 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Intelligent Vehicles
- Vol. 4 (2) , 287-297
Merging in Congested Freeway Traffic Using Multipolicy Decision Making and Passive Actor-Critic Learning
- 20 March 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Intelligent Vehicles
- Vol. 4 (2) , 287-297
Merging in Congested Freeway Traffic Using Multipolicy Decision Making and Passive Actor-Critic Learning