コーディング

Research

System Safety with AI Interface

 

 

Maintenance index

Maintenance index: A maintenance index has been developed to quantitatively evaluate the reliability of the safety functions of a nuclear power plant in terms of the probability of loss of function, taking into account the effects of both aging and maintenance activities.

Resilience index

Accident Management (AM) scenarios for the temporary loss of critical safety functions are developed, and their ability to recover safety functions, which cannot be evaluated by PRA, is quantitatively evaluated using the AM success probability, safety margins, and time margins as indices.

AI-based non-destructive testing (NDT)

The deep learning model is applied to nondestructive testing measurement signals to perform robust flaw depth identification to noise such as lift-off.

AI-based early failure detection

The coupling of time-series deep learning and statistical methods is used for early detection of abnormal conditions in plant equipment at a predictive stage.

AI-based action monitoring

Development of a primary screening system that automatically detects the appropriateness of human activities and crimes related to nuclear power, such as accident AM activities and nuclear security, through the interface between image AI and language processing AI.

オリジナル動画

再構成動画

Medical imaging application
(tumor dynamics analysis)

Predict the future of three-dimensional movies of lung tumors moving with respiration to realize tracking radiotherapy.

Access

Demachi laboratory

Department of Nuclear Engineering and Management

The University of Tokyo

7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656, JAPAN

Professor's room: Room 812, Engineering building 8

Students' room: Room 802, Engineering building 8

Tel:03-5841-8630

​​​​© 2020 by Demachi laboratory. Proudly created with Wix.com

  • Grey Facebook Icon
  • Grey Twitter Icon
  • Grey LinkedIn Icon