Teppei Mori
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Teppei Mori has a total of 28 co-authors. Here are the top co-authors by number of shared publications:
- Keita Ishii
Bridgestone Corporation,Digital AI/IoT Planning and Development Division,Tokyo,Japan
3 shared publications - Shunsuke Ono
Department of Computer Science, Institute of Science Tokyo, Yokohama, Japan
2 shared publications Takeshi Masago
Digital Solution AI IoT Planning and Development Department, Bridgestone Corporation, Tokyo, Japan
2 shared publications- Aki Shigesawa
Graduate School of Engineering, Hokkaido University, Sapporo 060-0808, Japan
1 shared publication Akira Saida
Civil Engineering Research Institute for Cold Region,Traffic Engineering Research Team,Sapporo,Japan
1 shared publication- Haruaki Yanagisawa
Department of Cell Biology and Anatomy, Graduate School of Medicine, The University of Tokyo, Tokyo 113-0033, Japan
1 shared publication
Latest Publications
- Open Access
Winter Road Surface Condition Recognition in Snowy Regions Based on Image-to-Image Translation
In snowy regions, road surface conditions change due to snowfall or ice formation in winter. This can lead to very dangerous situations when driving a car. Therefore, recognizing road surface conditions is important for both drivers and road managers. Road surface classification using in-vehicle cameras faces challenges due to the diverse environments in which vehicles operate. It is difficult to build a single classification model that can handle all conditions. One major challenge is illumination. During dusk, it changes rapidly and drastically, resulting in poor classification accuracy. Therefore, a robust method is needed to accurately recognize road conditions at all times. In this study, we used an image translation method to standardize illumination conditions. Next, we extracted features from both the translated images and the original images using MobileNet. Finally, we integrated these features using Late Fusion with an Extreme Learning Machine to classify road conditions. The effectiveness of this method was verified using a dataset of in-vehicle camera images. The results showed that the accuracy of this method achieved 78% during dusk and outperformed the comparison methods. It was confirmed that the uniformity of illumination conditions contributed to the improvement in classification accuracy. The proposed method can classify road conditions even during dusk, when sudden changes in illumination occur. This demonstrates the potential to realize a robust road condition recognition method that contributes to improved driver safety and efficient road management.
Open AccessWinter Road Surface Condition Recognition in Snowy Regions Based on Image-to-Image Translation
In snowy regions, road surface conditions change due to snowfall or ice formation in winter. This can lead to very dangerous situations when driving a car. Therefore, recognizing road surface conditions is important for both drivers and road managers. Road surface classification using in-vehicle cameras faces challenges due to the diverse environments in which vehicles operate. It is difficult to build a single classification model that can handle all conditions. One major challenge is illumination. During dusk, it changes rapidly and drastically, resulting in poor classification accuracy. Therefore, a robust method is needed to accurately recognize road conditions at all times. In this study, we used an image translation method to standardize illumination conditions. Next, we extracted features from both the translated images and the original images using MobileNet. Finally, we integrated these features using Late Fusion with an Extreme Learning Machine to classify road conditions. The effectiveness of this method was verified using a dataset of in-vehicle camera images. The results showed that the accuracy of this method achieved 78% during dusk and outperformed the comparison methods. It was confirmed that the uniformity of illumination conditions contributed to the improvement in classification accuracy. The proposed method can classify road conditions even during dusk, when sudden changes in illumination occur. This demonstrates the potential to realize a robust road condition recognition method that contributes to improved driver safety and efficient road management.
- 18 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 2173-2180
High-Resolution Forecasting of Road Surface Temperature Using Graph Neural Networks
Effective winter road maintenance (WRM) is critical for the safety and mobility of road users in cold regions. High spatial resolution predictions of road surface conditions greatly improve WRM efficiency. Previous studies have explored spatial modeling to achieve this; however, traditional statistical models, such as Kriging, face limitations in capturing complex spatial dependency. In this study, we utilized graph neural networks (GNNs) for spatial modeling to predict road surface temperature (RST), which an essential road weather parameter, at high spatial resolution. Furthermore, by integrating this approach with time-series prediction model at road weather observatories and weather mesh forecast data, we propose a framework that enables high spatial-resolution RST predictions over future horizons. Experiments using real-world data from Sapporo, Japan demonstrated that the proposed method could predict RST at approximately $\mathbf{1 0 0}$-meter intervals along road networks with an root mean square error below 2° C for future horizons up to $\mathbf{2 4 h}$. In conclusion, the study findings represent a significant improvement in accuracy over current operational methods, offering valuable support for decision-making in WRM planning to enhance road safety and mobility.- 18 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 2173-2180
High-Resolution Forecasting of Road Surface Temperature Using Graph Neural Networks
Effective winter road maintenance (WRM) is critical for the safety and mobility of road users in cold regions. High spatial resolution predictions of road surface conditions greatly improve WRM efficiency. Previous studies have explored spatial modeling to achieve this; however, traditional statistical models, such as Kriging, face limitations in capturing complex spatial dependency. In this study, we utilized graph neural networks (GNNs) for spatial modeling to predict road surface temperature (RST), which an essential road weather parameter, at high spatial resolution. Furthermore, by integrating this approach with time-series prediction model at road weather observatories and weather mesh forecast data, we propose a framework that enables high spatial-resolution RST predictions over future horizons. Experiments using real-world data from Sapporo, Japan demonstrated that the proposed method could predict RST at approximately $\mathbf{1 0 0}$-meter intervals along road networks with an root mean square error below 2° C for future horizons up to $\mathbf{2 4 h}$. In conclusion, the study findings represent a significant improvement in accuracy over current operational methods, offering valuable support for decision-making in WRM planning to enhance road safety and mobility.- Open Access
- 13 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Intelligent Vehicles
- Vol. 11 (1) , 185-196
Longitudinal Force Estimation in Intelligent Tires Using Key Features and Tread Dynamics Validation
Intelligent tire systems have garnered considerable interest as a technology that enhances tire safety through the monitoring of tire characteristics and tire–road interactions, which directly influence vehicle dynamics. However, battery life limitations owing to computational demands constrain their practical implementation. This study focused on studless winter tires to improve safety on icy and snowy roads, where freezing and snow accumulation increase braking distances, elevating the risk of accidents. Specifically, we developed computationally efficient features from tire acceleration signals to estimate longitudinal force, a key factor in tire–road interaction. Acceleration signals were analyzed to extract features most effective for force estimation. To reduce power consumption, only the most relevant features were selected. The selected features were applied to a machine learning model (ExtraTree regressor) to estimate longitudinal force. The method achieved high estimation accuracy with a normalized root mean square error (NRMSE) of 3.3%, while significantly minimizing computational load and power consumption. Compared to transmitting raw signals, the proposed approach reduced power consumption from 49.4 mW to 0.11 mW per second. Direct observations of the tire–road contact patch using a high-speed camera were conducted to validate the features. Time–frequency analysis of acceleration signals further supported the features' effectiveness, revealing that they correspond to tread vibrations caused by the relaxation phenomenon, where deformed tread elements recover after road contact. The proposed approach offers a promising method to enhance safety and efficiency in winter driving conditions by providing accurate, real-time tire–road interaction data while conserving energy.
- 13 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Intelligent Vehicles
- Vol. 11 (1) , 185-196
Open AccessLongitudinal Force Estimation in Intelligent Tires Using Key Features and Tread Dynamics Validation
Intelligent tire systems have garnered considerable interest as a technology that enhances tire safety through the monitoring of tire characteristics and tire–road interactions, which directly influence vehicle dynamics. However, battery life limitations owing to computational demands constrain their practical implementation. This study focused on studless winter tires to improve safety on icy and snowy roads, where freezing and snow accumulation increase braking distances, elevating the risk of accidents. Specifically, we developed computationally efficient features from tire acceleration signals to estimate longitudinal force, a key factor in tire–road interaction. Acceleration signals were analyzed to extract features most effective for force estimation. To reduce power consumption, only the most relevant features were selected. The selected features were applied to a machine learning model (ExtraTree regressor) to estimate longitudinal force. The method achieved high estimation accuracy with a normalized root mean square error (NRMSE) of 3.3%, while significantly minimizing computational load and power consumption. Compared to transmitting raw signals, the proposed approach reduced power consumption from 49.4 mW to 0.11 mW per second. Direct observations of the tire–road contact patch using a high-speed camera were conducted to validate the features. Time–frequency analysis of acceleration signals further supported the features' effectiveness, revealing that they correspond to tread vibrations caused by the relaxation phenomenon, where deformed tread elements recover after road contact. The proposed approach offers a promising method to enhance safety and efficiency in winter driving conditions by providing accurate, real-time tire–road interaction data while conserving energy.
- 29 April 2025
- Published by Rockefeller University Press in Journal of Cell Biology
- Vol. 224 (7)
Tension-induced suppression of allosteric conformational changes coordinates kinesin-1 stepping
Kinesin-1 walks along microtubules by alternating ATP hydrolysis and movement of its two motor domains (“head”). The detached head preferentially binds to the forward tubulin-binding site after ATP binds to the microtubule-bound head, but the mechanism preventing premature microtubule binding while the partner head awaits ATP remains unknown. Here, we examined the role of the neck linker, the segment connecting two heads, in this mechanism. Structural analyses of the nucleotide-free head revealed a bulge just ahead of the neck linker’s base, creating an asymmetric constraint on its mobility. While the neck linker can stretch freely backward, it must navigate around this bulge to extend forward. We hypothesized that increased neck linker tension suppresses premature binding of the tethered head, which was supported by molecular dynamics simulations and single-molecule fluorescence assays. These findings demonstrate a tension-dependent allosteric mechanism that coordinates the movement of two heads, where neck linker tension modulates the allosteric conformational changes rather than directly affecting the nucleotide state.
- 29 April 2025
- Published by Rockefeller University Press in Journal of Cell Biology
- Vol. 224 (7)
Tension-induced suppression of allosteric conformational changes coordinates kinesin-1 stepping
Kinesin-1 walks along microtubules by alternating ATP hydrolysis and movement of its two motor domains (“head”). The detached head preferentially binds to the forward tubulin-binding site after ATP binds to the microtubule-bound head, but the mechanism preventing premature microtubule binding while the partner head awaits ATP remains unknown. Here, we examined the role of the neck linker, the segment connecting two heads, in this mechanism. Structural analyses of the nucleotide-free head revealed a bulge just ahead of the neck linker’s base, creating an asymmetric constraint on its mobility. While the neck linker can stretch freely backward, it must navigate around this bulge to extend forward. We hypothesized that increased neck linker tension suppresses premature binding of the tethered head, which was supported by molecular dynamics simulations and single-molecule fluorescence assays. These findings demonstrate a tension-dependent allosteric mechanism that coordinates the movement of two heads, where neck linker tension modulates the allosteric conformational changes rather than directly affecting the nucleotide state.
- Open Access
- 5 August 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Intelligent Transportation Systems
- Vol. 25 (11) , 17122-17133
Fine-Scaled Predictive Modeling of Road Surface Conditions and Temperature in Urban Areas
Road administrators require fine-scaled information regarding road surface conditions to ensure efficient operation during winter periods. However, conventional models offer low-resolution information at a scale comparable to meteorological meshes or the spatial configuration of road weather information systems. Additionally, few methods have been proposed for predicting road surface conditions specifically in urban areas, where roads frequently experience shading from surrounding buildings. This study proposes a statistical approach for predicting road surface temperature and conditions in urban road networks. The complicated accumulated distribution of solar radiation along each road is calculated and used as an effective explanatory variable that considers the complex shading effects of nearby structures. The proposed model adopts a Bayesian spatiotemporal hierarchical framework for predicting road surface temperature using a solar radiation variable. Furthermore, a spatial machine learning model is implemented to estimate road surface conditions. The model classifies icy road conditions into six distinct types, achieving a sensitivity of 0.7712 and a balanced accuracy of 0.8637. Ultimately, the model provides significant information required for decision-making processes aimed at ensuring efficient winter road management. These results indicate that the applicability of the proposed approach can extend beyond the studied area, demonstrating its potential for broader implementation.
- 5 August 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Intelligent Transportation Systems
- Vol. 25 (11) , 17122-17133
Open AccessFine-Scaled Predictive Modeling of Road Surface Conditions and Temperature in Urban Areas
Road administrators require fine-scaled information regarding road surface conditions to ensure efficient operation during winter periods. However, conventional models offer low-resolution information at a scale comparable to meteorological meshes or the spatial configuration of road weather information systems. Additionally, few methods have been proposed for predicting road surface conditions specifically in urban areas, where roads frequently experience shading from surrounding buildings. This study proposes a statistical approach for predicting road surface temperature and conditions in urban road networks. The complicated accumulated distribution of solar radiation along each road is calculated and used as an effective explanatory variable that considers the complex shading effects of nearby structures. The proposed model adopts a Bayesian spatiotemporal hierarchical framework for predicting road surface temperature using a solar radiation variable. Furthermore, a spatial machine learning model is implemented to estimate road surface conditions. The model classifies icy road conditions into six distinct types, achieving a sensitivity of 0.7712 and a balanced accuracy of 0.8637. Ultimately, the model provides significant information required for decision-making processes aimed at ensuring efficient winter road management. These results indicate that the applicability of the proposed approach can extend beyond the studied area, demonstrating its potential for broader implementation.