Research | 東京大学　中野研究室

Overview

Mobility Measurement and Controlling

Based on the fundamental fields of mechanics, vibration, and control engineering, researches on state monitoring, ergonomics, human-machine interface, automated driving, and cooperative systems related to mobility are conducted. Non-technical issues, called ELSI, are also being addressed, with the aim of implementing these technologies in society.

Major Research Themes (click for details)

Evaluation of Performance of Shared Control: Shared control is a system controlling something cooperating with a human. A part of advanced driver assist systems of automobiles are corresponding to it. Our laboratory is conducting researches on a haptic steering guidance system as an example of the shared control.
　
Evaluation of Human Machine Interface for Vehicle-infrastructure Cooperative Driver Assistance: In the future, the social implementation of Level 4 automated driving is expected to begin with commercial vehicles, while Level 2 driver assistance systems are likely to expand to public roads for private vehicles. To facilitate Level 4 deployment, infrastructure-cooperative systems are being developed to transmit information such as traffic signals and detected objects from roadside sensors to vehicles. To extend the use of such systems to driver assistance, effective human-machine interfaces (HMIs) are being investigated through driving simulator experiments.
HMI Based on Human Foresight Prediction for Advisory Take-over Timing: For SAE Level 2 driving automation systems, a key challenge for Human-Machine Interface (HMI) is to help the driver re-engage in the driving loop quickly and safely. This study proposes an advisory HMI based on appropriate Pre-Timing, integrating scenario urgency and driver gaze behavior to enable safe and efficient driving take-over.
Understanding and Optimizing Situational Acceptance in Automated Driving: With the rapid advancement of automation and electrification technologies, low technology acceptance—driven by users’ psychological resistance—has become a major barrier to the commercialization of intelligent transportation. Meanwhile, most existing studies conceptualize acceptance as a pre-use decision to adopt or reject a technology, overlooking its dynamic and situational attributes during interaction. To address this gap, a series of driving simulator studies were conducted
Toward Interpretable Urban Automated Driving through Style-Aware Trajectory Planning and Semantic Reasoning: This research proposes a system-level interpretable framework for automated driving, covering perception, planning, and decision-making. It aims to enhance transparency and human-aligned interpretability across all levels.
Unified Traffic Control System for Railway and Road Vehicles Using Mobile Phone Line: Unified traffic control system for railway and road vehicles using mobile phone line is proposed. The demonstration is carried out at ITS experimental field in Kashiwa campus, including the railway test track, the test road, and the railway crossings.
Enhancing Multi-Agent Traffic Signal Control in Complex Environments Using Large Language Models: Existing traffic management systems often fail to effectively respond in emergency scenarios. When some long-tail events occur, manual intervention by traffic wardens is typically required. Given the human-level reasoning capability of large language models (LLMs), they show strong potential for solving complex tasks.
Tram Vehicle Clearance Identification Algorithm To Enhance Tram Operational Safety in Mixed-Traffic Environments: To improve tram operational safety, it is undeniably crucial to equip trams with the capability to perceive the surrounding environment, especially the front area, to prevent collision accidents. Thus, the main purpose of this study is to introduce an algorithm that identifies tram vehicle clearance, defined as the distance from the rail track center to both ends of the tram vehicle. This clearance identification helps determine whether any obstacles are within the tram’s path.
Machine Learning-Based Estimation of Condition Between Rail and Wheel from Measured Values of a PQ Wheelset: Running safety against flange-climb derailment is evaluated using the derailment coefficient Q/P, defined as the ratio of lateral force Q to wheel load P. However, even for the same derailment coefficient, the likelihood of derailment varies depending on the wheel–rail contact conditions. Therefore, it is necessary to establish safety limit criteria that take into account the contact state between the wheel and rail.In this study, a method is proposed to estimate the angle of attack, lateral contact position, and friction coefficient using machine learning based on measurement data obtained from a PQ wheelset that measures the derailment coefficient. The validity of the proposed method is then examined through numerical simulations.
Activities to Realize Level 4 Cooperated Automated Mobility Service: Pilot test to run an automated driving bus (Level 2 operation) between Kashiwanoha Campus Station and Kashiwa campus of the University of Tokyo every day (weekdays only) started in November 2019 implemented by Kashiwa ITS Promotion Council. To link these activities to the social implementation of Level 4 automated driving services in the Kashiwanoha area of Kashiwa City, a six-party consortium led by the University of Tokyo was entrusted with “RoAD to the L4 project (Theme 4)”, called as CooL4, of the Ministry of Economy, Trade and Industry and the Ministry of Land, Infrastructure and Transport.
Building the Method for Social Implementation of Automated Driving Technology Complying with Actual State Based on ELSI: ELSI stands for Ethical, Legal and Social Implications/Issues. There have been efforts in various fields to study and deal with ELSIs that arise with the development of new science and technology. This study examined how automated driving technology should be implemented in society, based on the fundamental question, "Can humans and society accept the mistakes made by machines?”. This project was supported by the Research and Development Center for Social Technology (RISTEX) of the Japan Science and Technology Agency (JST) under the "Research and Development Program for Comprehensive Application of Science and Technology to Ethical, Legal, and Social Issues (ELSI)" (RInCA) (FY 2020), and was conducted by the University of Tokyo, Meiji University, and the University of Tsukuba. The project was conducted by the University of Tokyo, Meiji University, and University of Tsukuba.
Effects of Guidance Intensity in Driving Assistance HMI on Driving Behavior: Although driving assistance technologies have advanced, not all forms of assistance necessarily improve the safety of driving behavior. When humans receive information intended to prompt an action, precise information may lead to lower performance compared to ambiguous information. However, there are few studies that have examined whether this finding can be applied to automobile driving or driver assistance systems. In this study, we aim to investigate the effects of guidance intensity when presenting information on safety driving behavior.
Vehicle Dynamics and Safety Assessment of Railway Train Set during Earthquakes: Large earthquakes repeatedly occurred in Japan, and in some cases, derailment and overturning of railway vehicles occurred during these earthquakes. Safety measures are being implemented in various railway sectors, and it is important to continue studying to improve safety. On the other hand, it is difficult to conduct experiments involving derailment and overturn using actual vehicles. Also, conventional vehicle dynamics simulation VDS can mainly handle vehicle behaviour up to derailment. This study is conducting research to overcome these problems and to understand the derailment and overturn phenomena by using model vehicles and developing a new vehicle dynamics simulation VDS-X.

Experiment Device