Skip to main content

Week 04

Jan 23

  • create flow diagram
  • train models using data where ND distribution is about the same as other defects
    • randomly delete row with ND
  • Create table that showing the result of different run configuration
  • create a model using restnet 34, pretrained = False
  • Label on the SD1 images and manually set aside images for testing

Jan 24

  • set 15 images each video that have not been labelled SD1
  • train using the 3 videos and train using all the videos (keep the same testing image)
  • labels all the images

Jan 26

  • Calculate the azure account. How many images can we label (item)
  • prediction of the new DNV videos (item)
  • items issue with access database and label images (item)
  • Send the FME
  • Document my thought on what I want I want to do with the new DNV videos: https://people.gqc.com/people/Deven/Daily%20Updates/2023/January/Week%204#new-data-from-dnv
  • Send email to sean asking for access database corresponding to the videos
  • After looking at the video the defect annotated in the video does not follow the current PACP standard. Asking sean what standard did they follow?

Jan 27

  • email to Sean.

  • make sewerml run again with new script

  • SD1

  • list of question to Brian

  • FME

  • Questions to Brian:

    • Call model defined parameter from python script block
    • In notebook,how do I change the
    • Convert

Meeting with Brian

  1. List of requested features a. Ability to read model defined parameter within the python script block (in progress) b. Ability to update submodel parameter from the notebook using collimator python-api (in progress) c. Provide a menu option or script to convert old submodel block to the current submodel block     d. Ability to handle algebraic loop problem (the unit delay block uses global step and does not address the problem)     e. Ability to search for a model/file in the project folder     f. A way to organize different files present inside the project folder (subfolders?) (in progress)       g. Ability to save runs? If a simul takes 3 hrs to finish, and you forget to save the results, will collimator have the ability to save those results? (they currently have a box icon at the top right corner where you can see the results from the last three days? or three hours simulation.)
  1. Marketplace mechanism, if GQC creates custom blocks inside Collimator, what would the licensing mechanism be? Where would those blocks get hosted?       a. How do we encrypt custom blocks?
  2. Share some updates about the publication
  3. Share the results from GQC collimator's workspace
  4. We want to have a shiny interface to the Collimator model and results that would be similar to the SCADA HMI. How would we go about doing that?
  5. Any discrete block will follow the time simulation using global discrete clock time.
  6. Interpolation for datasource reader block:
    1. Zero order hold is used if we know that sensor reads the data at a given time interval
    2. Linear can be used if the we know that sensor is continuously giving us the data without a specific time step