Forum Discussion
LouisWambsganss
6 years agoQrew Trainee
Evan,
Thanks for the reply. The tables are currently separate. They are populated from separate .csv files. How would I build the necessary table relationships?
------------------------------
Louis Wambsganss
------------------------------
Thanks for the reply. The tables are currently separate. They are populated from separate .csv files. How would I build the necessary table relationships?
------------------------------
Louis Wambsganss
------------------------------
EvanMartinez
6 years agoModerator
Hi Lou,
If you were creating a relationship would it be the case that one Discrepancy can have many Parts but any one Part would only ever be related to one Discrepancy? Or is it possible that a single part record might tie back to multiple Discrepancies? Also is there any piece of data that appears on both that would be unique? Like an incident number or anything?
------------------------------
Evan Martinez
Community Marketing Manager
Quick Base
------------------------------
If you were creating a relationship would it be the case that one Discrepancy can have many Parts but any one Part would only ever be related to one Discrepancy? Or is it possible that a single part record might tie back to multiple Discrepancies? Also is there any piece of data that appears on both that would be unique? Like an incident number or anything?
------------------------------
Evan Martinez
Community Marketing Manager
Quick Base
------------------------------
- LouisWambsganss6 years agoQrew TraineeEvan,
Again, thanks for your time. It is possible that one Part has multiple discrepancies, but one discrepancy will only affect one part.
In my context, a Machine is a complex assembly made up of many parts. Each part is tracked by a Part # and Serial #. A Machine is constantly having parts removed and installed.
The issue I have is that I cannot automate the assignment of a specific part (Part # and unique Serial #) to a discrepancy. Due to variables in how technicians describe issues, I cannot rely on automated sorting or filtering. A technician will create a discrepancy that just says "Machine #37 is broken". So I need to have humans read each discrepancy text and decide "OK, this is a problem with a Main Widget" or "This is a problem with a Secondary Pressure Switch" etc. as applicable.
In order to help streamline this process, I'm looking for a way to link Part Installation data onto a form in the Discrepancy Table. My thought is that when a user views a Discrepancy, the form will take data from the Discrepancy Table (Discrepancy #, Machine #, and Discrepancy Date) and will display all Parts where Discrepancy Date falls between Part Installed Date and Part Removed Date, and Discrepancy Machine # equals Part Installed On Machine # . So the User will be given a list of all parts that were on that machine on that date, and they can choose the correct part.
Here are two example Discrepancies from the Discrepancy Table. I need to fill in the Part and Serial fields.
Discrepancy Table Discrepancy Machine Discrepancy Date Description Discrepancy Part Discrepancy Serial 1100 45 20-Jan-05 Widget broken ??? ??? 1540 37 1-Aug-14 Not working right ??? ???
Here are two sample entries from the Parts Table:
Parts Table Part # Serial Number Installed Date Installed on Machine Removed Date 12498 357A 12-Feb-04 45 16-Sep-09 25497 498B 23-Apr-19 37 (NULL, still installed)
In this case, I can see that Part 12498 (Serial 357A) was installed on Machine 45 at the time when Discrepancy 1100 was raised, so I would choose this part to fill in Discrepancy Part and Discrepancy Serial. I just need to be able to do this for 15,500 Discrepancy records, pulling from a Parts Table with nearly 1 Million records. Due to the large quantities, I'm trying to automate as much of the research as possible. Ideally, the human effort in this would be minimized to just interpreting "system hydraulic pressure fluctuates" and selecting "pump A", rather than manually researching what pump was on that machine at that date.
------------------------------
Louis Wambsganss
------------------------------- MarkShnier__You6 years ago
Qrew Legend
Louis
I do know how to do this.
My thought is that when a user views a Discrepancy, the form will take data from the Discrepancy Table (Discrepancy #, Machine #, and Discrepancy Date) and will display all Parts where Discrepancy Date falls between Part Installed Date and Part Removed Date, and Discrepancy Machine # equals Part Installed On Machine # . So the User will be given a list of all parts that were on that machine on that date, and they can choose the correct part.
I can help you one on one to implement a solution, but here would be the proposed workflow it this meets your need.
The user views a Discrepancy and on that record click a button. The button will edit or add to a table of users called User Focus table where the key field is the Userids. The particulars of the Descrepancy record will looked up by a reverse relationship to the User Focus record.
Then those fields will be looked up down to ALL Parts and Parts which meet some complex criteria will be displayed immediately on the Descrepancy record. The user can then click a button on the part record to select it whatever that means - presumably it means to add a record on a table somewhere or to edit the Descrepancy record to to assign the cause of a problem to a Part. That same button could also redirect to the original list of 15,500 Discrepancies to be coded, however now the report will "only" be 15,499 records long and you just put a bunch of staff onto it all hammerin' away until you beat that list down to zero records.
So the goal is to minimize the clicks and make every one count since you are saying that a Human needs to have a quick look to interpret the comment and then assign the blame to a part.
It's probably just an hour or two of one on one consulting time required.
mark.shnier@gmail.com www.QuickBaseCoach.com
------------------------------
Mark Shnier (YQC)
Quick Base Solution Provider
Your Quick Base Coach
http://QuickBaseCoach.com
markshnier2@gmail.com
------------------------------ - EvanMartinez6 years agoModeratorHi Lou,
Yeah I would have to say that is a pretty tough workflow. Ideally you would have those two records be directly related so you could pass down things like a date range form the parent records to the child record and then set your reports to filter off of that. Unfortunately at this instance it doesn't sound like you have a unique consistent value between the two tables to set up your relationship off of since in order to include something in a report/filter it needs to exist in some way in the table you are working on. My suggestion would be to put in a case with the Care team and walk through what you are looking to do in a screen share just incase they can see something I am missing from this description, sometimes getting to see the app live and walk through the workflow needed can help it all gel together but that sounds like a pretty complex one.
------------------------------
Evan Martinez
Community Marketing Manager
Quick Base
------------------------------