That would be useful, however expiration dates are often printed so badly and in all kinds of locations that it seems difficult to make that work reliably. Also people wouldn’t want photos of their fridge contents being uploaded to a cloud AI all the time.

I have the solution for those labels, even if that hype is over for quite a while now: RFID tags! /s

Can somebody enlighten me why the above statement was sarcastic and why it was downvoted?

Aren't RFID tags on all groceries a viable solution to make product info available to a fridge?

Because RFID labels are orders of magnitide more expensive than bar codes and printed labels. The groccery margins are simply to small for that. Hence the sarcastic comment of using expensive tech (that had its own hype curve in the early 2000s), to solve the how-do-I-know-whats-in-my-fridge-without-opening-it problem.

IF RFID was useful for other purposes it would be worth the small cost (scale can bring RFID costs down - it would cost more than bar codes, but potentially can bring enough logistical benefits that overall it is cheaper). However so far the logistical benefits can't be realized because you can't be 100% sure you read all the tags in a room.

Mass readability is one of the reasons why logistics isn't using RFID tags atva global scale. Cost is another, and a major one.

Sounds like a good idea until I think about our fridge. After the weekly shop the fridge is packed to the brim so much that the light barely makes it into some corners. How would a camera help here? Towards the end of the week the fridge is empty and I know I need to buy the usual. Use case dead in my head.

It would be nice if there was a standardized machine-readable format for information like expiration dates. It would be neat to be able to track inventory in the fridge by just scanning a label when you put it in then scan and mark as empty when its done.