21–30 of 152 posts
One thought to reduce data entry (for some people): parse freshdirect (etc) emails, with a default expiration date put in. (The pie in the sky version uses machine learning to make better guesses.)
I love the idea of this, but in reality I can’t imagine the time and energy required to scan and keep everything up-to-date is offset by the benefits. Perhaps a current user can prove me wrong?
I wrote something very similar to this a couple years ago for our household. It runs on a raspberry pi and uses a barcode scanner and when we go shopping we scan everything in and then scan things out as we use them. We live far away from town and shopping trips are rare, big all day events. It's easy to forget to scan things out and then things that don't have barcodes like vegetables are hard to track so we don't bother. Manually entering it all in and then remembering to remove it while cooking is too much trouble (we tried).
(Personally working with "out of home" logistic solutions.)
Edit: mistypes.
I love the idea of this, but in reality I can’t imagine the time and energy required to scan and keep everything up-to-date is offset by the benefits. Perhaps a current user can prove me wrong?
Agreed. If there was some kind of tech like Amazon's just walk out idea that could automate keeping track of everything it would be great. But having to do this stuff manually kills it. It's the same problem I have with personal money mangers; I have to manually import my bank transactions and then further manually categorize and label them (I've scripted some but can't do most) for it to properly work and give me go…
One of the options to solve the data entry could be: - good quality automatic scanning of receipts (not only individual barcodes) from shops using OCR possibly supported with image recognition for double-checking (can happen that products will be mis-labelled or without quantities, etc) - when ordering on-line, the receipt should be available, so should be also much easier Yet, not always one will have a meaningful receipt available...
Solved the data entry and being able to predict own's supply needs would be also great to have a up-to date management of the inventory. Here are even more challenges on the tracking of the available goods at home, where these are and how many items (and in what state, expiration date, etc) would require most probably implementing different solutions from IoT (connected cameras, sensors, etc.).
Then, having a connected home with own groceries supplies under control, one can then automate further the shopping process with feeding-back the information about own's demand to on-line groceries one is subscribed to. This can enable customer subscription plans, and for retailer keeping a possible continuous flow of goods. This could be really really useful especially for upcoming months, when it seems like we are expected to spend a bit more time at home rather than usual, hopefully not fighting in the local shops for the last rolls of the new white paper gold.
I love the idea of this, but in reality I can’t imagine the time and energy required to scan and keep everything up-to-date is offset by the benefits. Perhaps a current user can prove me wrong?
I think this sort of inventory system combined with automatic updates (a la Amazon Go store style OCR embedded in smart fridges, etc) would go a long way. Have always thought this would be useful, but agreed the amount of energy required to manually update everything would be unsustainable.
Meats and vegetables would still be a challenge though, but perhaps object recognition could help with that.
Fridges are already overpriced for what even the fancy ones can offer, I imagine we're not there yet to incorporate multi-camera vision systems with beefy AI processors at commodity prices yet.
Called “standard pantry”
This is exactly what I needed and was thinking about building a similar app. But to be really useful and automated, there's a bit long way I'm afraid. The biggest deal breaker is the time spend for the data entry and updates, leaving a pen-and-paper / notes solution still the most easy to use and just visiting local groceries and shops. I was collecting receipts and playing a bit with scanning of these, but the amoun…
The other big challenge I never resolved was how you'd account for e.g. using 1/4 cup of flour out of a bigger volume. Or taken to its extreme, cooking oil. How do you know how much your 'splash' is? You can't predict the remaining volume without a lot of fiddling to measure it and that defeats the purpose.
In the end, I opted for manual databases too, but they're pain to keep up-to-date. I still think there's a lot of value in a database for all-or-nothing style ingredients, but it was enough to deter me. I'm glad someone is less lazy.