Believe me it works, I tested it, I had a database with 121 entries in the PrmMilkVisit table of all 0's, I simply entered a dry off date for 1 animal and presto, there were no more entries in the PrmMilkVisit table with all 0's.

This is a quick an dirty fix for something the real programmers wont fix for us coming from the robot's embedded software database.... for some odd reason random events get created in the PrmMilkVisit table where all the coordinates are 0 when they should be numbers (X,Y,Z robotic coordinates).

My fix is simply replace these 0's with a calculated average for each column..... but since these are random and the issue lies when the farmer enters a dry off date and the last milking for this cow just happens to be all 0's then the next time he enters the calving date, the robot doesn't store the coordinates properly for up to 7 milkings....when it should right away to automate the second visit to the robot, instead he has to position the arm up to 7 different visits to get them automated...very frustrating.

I simulated manually entering values on a problem cow and the second time in the robot she milked immediately without asking for manually setting the arm again.

So I came up with this idea that these 0 anomalies can be fixed every time the farmer is entering in new cows or calving dates and dry off dates...

It works like a charm....