Yoshifumi Nishida

Prime Senior Research Sientist, Artificial Intelligence Research Center
National Institute of Advanced Industrial Science and Technology (AIST)
E-Mail: y.nishida@aist.go.jp

Mar. 1998,	Completed Dr.s' course, Graduate School of Mechanical Engineering, The University of Tokyo.
Mar. 1998,	Received Dr. of Engineering from The University of Tokyo.
Apr. 1998	Joined Electrotechnical Laboratory, Agency of Industrial Science and Technology, MITI.
Apr. 2001,	Joined Digital Human Laboratory(DHL), National Institute of Advanced Industrial Science and Technology (AIST).
Apr. 2003	The laboratory was reorganized to Digital Human Research Center(DHRC), AIST.
Apr. 2003	Human Behavior Understanding Team Leader, DHRC, AIST.
Oct. 2005	Senior Research Scientist, DHRC, AIST
Oct. 2010	Injury Prevention Engineering Team, DHRC, AIST.
Oct. 2010	Leader of Team for Desigining Everyday Life System Function and Social System, DHRC, AIST
Oct. 2011	Prime Senior Research Sientist, DHRC, AIST
Oct. 2015 to present	Human Informatics Research Institute, AIST
Oct. 2015 to present	Prime Senior Research Sientist, Artificial Intelligence Research Center, AIST
Oct. 2016 to present	Leader of Living Intelligence Research Team, Artificial Intelligence Research Center, AIST

• Active Understanding Human Intention through Monitoring Human Behavior (1994-1995) as Master Work
  
  Understanding human intention is an essential function for a robot which can offer adequate support to human beings. It requires smooth communication between the human and the robot. Human behavior is an expressive media of communication. This work proposes a new function of "active understanding of human intentions" by a robot through monitoring of human behavior. The unique feature of the proposed function lies in the fact that it utilizes multi-communication channels in parallel, i.e., human intentions are understood not only through conscious behavior but also through unconscious behavior. This work also proposes a robot architecture to realize the function.Following points are the key features of the architecture: 1) A robot possesses multi-sensors which surround the human. 2) Information processing is carried out by dual loops a loop for information exchange between the human and the robot and a loop for human intention understanding. As an example of a robot with the human intention understanding functions, the authors constructed a micro-teleoperation robot. It can automatically understand an operator's intention through such unconscious behavior of touching a desk with an operator's hand which holds a pen-shaped master. The understood result is utilized to change the control mode of a master-slave manipulation system from fine motion to rough motion and vice versa. The experimental results prove that the proposed function is effective in making the operation of the system easier. Consequently the system is friendly to the operators.
  
  Application of a Function of Active Understanding of Human Intention to Micro-teleoperation System

  
  

• Understanding Human Physiological Status by Surrounding Sensor System (1996-1998) as Doctor Work
  
  This work develops a new method of understanding human physiological status through unrestrainedly and noninvasively monitoring his or her body movement by a surrounding sensor system. The new understanding method consists of the function of inferring physiological values (ex. tidal volume) from sensor signal , and the function of continuously monitoring of the sensor signal relating to target physiological values. This works proposes a "hierarchical human model" which consists of physiological layer and physical layer to realize the above functions respectively. The developed surrounding sensor system is composed of cameras embedded in a ceiling and 221 pressure sensors attached to a bed. Based on the surrounding sensor system and the hierarchical human model, a new method for diagnosing Sleep Apnea Syndrome(SAS) and monitoring posture in sleep is developed. Experiments to prove the effectiveness of the proposed method. were conducted not only in a laboratory but also at a hospital.
  • Surrounding Sensor System
    Development of Pressure Sensor Bed
  • New Method for Noninvasive and Unrestrained Understanding of Human Physiological Status
    • Noninvasive and Unrestrained Diagnosis of Sleep Apnea Syndrome by Image Processing
    • Noninvasive and Unrestrained Monitoring of Posture and Respiration by Pressure Sensor Bed

  
  

• Modeling and Monitoring Human Respiratory System Based on Environment Sensorization (1998-now)
  
  This work develops a living-space-shaped system for non-invasively and unrestrainedly monitoring the human respiratory system and intelligibly reporting its results. This environmental system consists of sensorized furniture: 1) a ceiling dome microphone, 2) a pressure sensor bed, and 3) a washstand display. The ceiling dome microphone consists of a ceiling dome, a lighting fixture, and an omnidirectional microphone. The ceiling dome microphone can detect not only snoring sounds but also normal breathing sounds, i.e., airflow at the mouth and nose. The pressure sensor bed has 210 tactile sensors which was developed by cooperating with a company based on the previously developed pressure sensor bed and can monitor body movement, breath curve and posture. By analyzing the breath curve, the system can estimate oxygen desaturation frequency. By integrating the above functions, for example, the system can find obstructive apnea, which is a typical apnea such that the patient cannot inhale the air despite of breath effort and that the concentration of the oxygen in blood falls. The washstand display can provide information related to the conditions of the respiratory system in the person's daily life. To prove the effectiveness of the integrated system and its functions, experiments are conducted for more than 30 real patients suffering from breath disorder.
  
  
  • Environment Sensorizing System
    • Improvement of Pressure Sensor Bed as Living Space-shaped Sensor
    • Development of Ceiling Dome Microphone as Living Space-shaped Sensor
    • Development of Washstand Display as Living-space-shaped Display
  • Digital Human Model
    • Comprehensive Model of the Human Respiratory System
  • New Method for Noninvasive and Unrestrained Monitoring of Human Respiratory System
    • Noninvasive and Unrestrained Monitoring of Blood Desaturation Frequency by Analyzing Breath Curve
    • Noninvasive and Unrestrained Monitoring of Respiratory Ventilation by Analyzing Breath Sound
  • Application
    • Sustainable Health Management Support by Health Information Presenting Furniture of Washstand Display

  
  

• Enabling Environment for Observing, Modeling and Supporting of Human Functions Based on Digital Human (2001-now)
  • Environment Sensorizing System
    • Development of Ultrasonic 3D Tag System for Object Sensorization
    • Minimally Privacy-Violative System for Observing Human Behavior
    • Method for Robust Observation, Recognition and Description of Daily Human Activities by Object Sensorization
    • Method for Integrating Distributed Sensors/Effectors
  • Digital Human Model
    • Infant Behavior Simulation
    • Model of Human Activities in Cognitive, Physiological, and Motion Level.
    • Method for Integrating Digital Human Components
    • Method for Integrating Digital Human Model and Environmental Sensor
  • Digital Human Presentation
    • Behavior-to-X engine