2. Human-machine cooperative mobility system

Human-machine cooperative mobility system

  • School of Engineering/Graduate School of Engineering
  • Department of Mechanical Science and Engineering
  • Subdepartment of Mechatronics

tatsuya suzuki [Professor]


Outline of Seeds

The analysis and modeling of human behavior is one of the key technologies to develop the human friendly system. The human behavior is investigated from the viewpoint of (continuous/discrete) hybrid dynamical systems. As a result, both of the decision making and motion control functions are modelled simultaneously. The driver-vehicle system and human-robot cooperative system are particularly focused on as typical applications. In addition, the assisting system is designed based on the identified behavior model and verified through some demonstrative experiments.

Novelty and originality of this research

In modeling behaviors, our team is able to obtain a physical understanding of relatively complex behaviors by creating a hybrid dynamic system (HDS) that, rather than modeling human behavior as a black box, creates relatively simple models of individual functions (human awareness, judgment, and actions) and then puts them together. By making extensive use of the system identification technologies in an HDS model like this, it is possible to obtain models from behavioral observation data without prior knowledge.
In addition, applying this HDS can help design HMI assessments, behavioral support systems that predict actions, or automated drive systems that conform to individual drivers.

Application and research area for Industry collaboration

-Driving behavior analysys based on deriver model
-Model based evaluatino of assitance system
-Model predictive driver assistance system
-Automated driving based on real time optimization and/or control technique
-Human in the loop simulation embedding driver model

Key Takeaway

We are participating in a variety of joint research projects with automotive companies where we use the technologies we have acquired through our behavioral models as a base and then combining them with optimization methods or control technologies. Please contact us if you are interested in learning more about our work.


Dynamic model, driver model, driving support, automated driving, optimization, driving evaluation


  • System identification of hybrid dynamical system
  • Design of human cooperative assistance system based on the control theory utilizing human behavior model
  • Automated driving system based on real time optimization
  • Model based evaluation of driving behavior for human machine interface or assistance system design


  • Small EV(as automated control testbed )
  • driving simulator
  • electric wheel chair
  • eye tracker
  • lidar sensors
  • electroencephalograph
  • force sensor, etc.


  • Japanese Patent Application No. 2015-65761
  • Japanese Unexamined Application Publication No. 2012-030659
  • Japanese Patent Application No. 2014-195206

Monographs, Papers and Articles

  • H. Okuda, et.al. “Modeling and Analysis of Driving Behavior based on Probability Weighted ARX Model,” IEEE Trans. on ITS, V14, No1, 2013
  • T. Yamaguchi, et.al., Implementation and verification of supervisory cooperative control by model predictive method, IEEE ITSC2016, 2016
  • H. Okuda, et.al. , Modeling and analysis of acceptability for merging vehicle at highway junction, IEEE ITSC2016, 2016
  • A. Koga, et.al., Realization of Different Driving Characteristics for Autonomous Drive by Using Model Predictive Control, IEEE IV2016, 2016