The task of empty shelves?

Question

IT education

An event either happens or it doesn't, and it all happens in time.

Let's say we have grocery stores. Each of them has the same products. What ways are there to determine if the shelf is empty or not?

It is necessary to determine, for example: is 40 minutes of no sales enough to say that this product is not on the shelf ...

What are the ways to implement this?

how to solve it statistically?, the chi-square came to mind, but I came to the conclusion that it does not fit here

Answer 1

no_name_0 ,

This is a grocery store
checks:
with time of purchase
with goods with quantity
with store number

You won’t believe it ...
It’s enough just to clearly fix the amount of goods put on the shelves or removed from them. Then the quantity of goods on the shelves based on this information and information from cash receipts can be COUNTED ! Well, plus or minus for broken, dragged into the next section and thrown anywhere, left at the checkout due to, say, lack of money, stolen, etc. All that matters is to screw up the merchandisers and set up normal accounting.
And you're here with the probabilities of foolishness toil.

Answer 2

Do you have enough probability that the shelf is empty? to get this information, you need to have reliable statistics on hand - sales history and the exact moment when the shelves became empty.
Shortcomings of the data (for example, their inaccuracy, or for example, there is no data on the exact history of sales, but there is their aggregation by day) will determine the algorithm. Also, the algorithm should work not only with these data, but also with others, for example, time of day, day of the week, month, upcoming holidays (both past and future) and so on. The more information, the fewer errors the algorithm will give.
Describe, in detail, all the information you have.