Ok just running an idea past you guys, so I ordered Amal’s book (which for some reason took a month from the seller to get to me) I was just trying to get more info in RFID and get some project ideas, it is a pretty cool book by the way. Anyway, one of the projects he mentioned is making a shelf with an antenna under it and using RFID tags to check what’s on the shelf. I was thinking it would be awesome to tag the stuff in my pantry and put in the antenna, then I can track what is in my pantry and I can stop buying pasta sauce every time I go to the store. I was thinking it could go a step further and make an android app, scan the barcode of the product, scan the RFID sticker, and put the item away. If you really wanted to get involved you could probably add purchase and use by dates too.
So the first question does this seem worth it as a project, I mentioned it to my wife and she didn’t seem very impressed. The second question is supplies, Amal mentioned the material he uses in the book, but the book is from 2006 so I figured there would be better options now. I’ve been trying to do some research on what board to use, Amal lists the Skytek M1 board, but I dont think they are in business anymore.
So what do you guys think, is it worth me spending some more time and energy on or not?
What you want to do (read a lot of tags with one antenna) entails:
Long(er) distances
Proper collision detection
Meaning, if you do your homework, you’ll find the only kind of RFID system that’ll work for you is UHF - and not the cheap stuff. Matter of fact, is there any cheap UHF?
In short, if you want a reliable system, it’ll cost you a bundle. Other than that, yeah it sounds interesting.
I got a long range reader that has anti collision and works well for just over 100 usd not sure what counts as cheep but I was happy with that. I’ll find the product page later.
Does it hit all the tags reliably 100% of the times in any tag position? The two sub-250 readers I bought out of China work well, but not well enough for exact inventory tracking. Just well enough for statistical tracking.
Mine is pointed out my window down my access pathway
it always picks up the tag during my testing (projects still a wip)
But that is me walking 4m through the field so like a lot of chances for a read. I will do some shorter range testing next time I work on the project see its reliability.
Cheap is a relitive term I’ve found out, to me cheap means it will work I won’t have to replace it anytime soon and I still have all of my limbs after paying for it. So 100 usd would easily fit in the catagory of keeping all my limbs lol.
This might be of interest to you, I think it would do what your are trying to achieve perfectly.
However
I think the one “issue” would be, the enrolling of every item in you pantry, then removing it (and the RFID sticker? that you can buy cheaply from China shopping sites) from your inventory and then attaching a new RFID sticker to every new item after shopping and enrolling it into your inventory.
Though it pains me to say it, I think you would be better suited looking at a barcode system, since manufacturers have already done the “sticker” part for you, and you can get an app like
and you can use your phone to read the Barcodes.
It all sounds like a “fun” project, but a little too much work for me, I would accept my fate of “too much” pasta sauce and think of it as an investment into food stocks for the upcoming apocalypse
Good luck, and let us know what path you take and how it goes…
well, I mean, both ISO14443 and ISO15693 support anti-collision… and most readers do too… it’s typically the application software/firmware where things fall down. The tag select process for ISO14443A for example can easily query all tags in the field and get back all the IDs in the field at once, and not even have to select any of them, meaning all tags remain active while in the field. If the reader wants to interact with a specific tag, then it issues a select to the ID it wants to talk to and then the other tags all shut up… but it’s not necessary for inventory applications like this where you just care about what IDs are in the field.
UHF is considered the inventory tracking tech of choice only because it typically has a wide field and much longer range, so objects moving through the backscatter field have many many chance and angles to be easily picked up (warehouses, conveyor belts, etc.)… however, there are some caveats… the most important one I think is that UHF performance is basically crap when tags are affixed to items that contain water (pasta sauce) and metal (tuna cans) so unless you spring for the expensive “on metal” UHF tags, it’s not necessarily going to have superior performance over HF. Granted, HF also has this same issue, but I’m what I’m saying is that UHF isn’t necessarily ideal for this particular project.
In short, UHF or HF both support anti-collision and can read hundreds of tags in any given field at once.
Personally, I think this project sounds like a key candidate for computer vision… let the computer watch what you put away and take out and get glimpses of every time you open the pantry. Put that ML/AI shit to work reading labels and identifying food items. Cameras are super cheap and no extra labor for your wife or family when putting shit away. When it comes to projects like this the point is to be more convenient, not less… and to me it feels like the management aspect of having to tag products and deal with all that is more ongoing management hassle than it’s worth. I liked books and other things for the project in my book because those items rarely move and once tagged, they are tagged for life… unlike food which cycles in and out constantly.
I’d say if you want to do this type of project, pick something else that doesn’t require constant management of the items to keep them tagged, updated, etc.
Tons of good points. I was thinking maybe using active tags that get reused for the same product. You don’t need to track one off things either. Unless you always have that item, it doesn’t make sense to track. It would be way less work to set it up for staple food items. Have a load cell under the floor, sugar, eggs etc that reports how much you have left. Maybe even use tags to see when it’s been removed from the pantry and track how much is being used day to day.
Now that I’m thinking, you wouldn’t even need tags, just the lack if weight could trigger the removal.
It definitely would be cool, but I feel like a system where you enroll groceries as they’re put away wouldn’t last too long in my world!
Thats why I was gonna use cheap sticker tags. If the system didn’t detect a tag I would know I was out of it. Thats the reason for an app too if the tag wasn’t found it could auto populate a shopping list. At least thats how it works in my head. Lol.
In my head, you would open your app, and "add canned tomatoes " and you would have stickers designed for only canned tomatoes. So from your app you could add 5, and put them away.
When it came time to use things, you would scan them like at a check out, and it would let the app know to let you know you are almost out.
But then it would become an issue of having stickers to use. I WOULD HATE to program each thing to its own label. I wohld use basics like, frozen meat, canned veggie, canned fruit, but that isn’t helpful much either if you need canned peas and only have carrots.
Ya, with all the stickers and scanning I would use it for a week or two and get annoyed and stop using it. However I’m definitely gonna use Amels suggestion and try to use my budding AI and programming skills, we’ll see how it goes.
that’s exactly the kind of thinking i’m talking about… if you had designated spots on the shelf for certain types of foods, you could easily use cheap load cells for each section. then again, my hunch is that if you were that organized as to have a place for everything and everything in its place, you probably would be acutely aware of what you did and didn’t have.
I’d say build it. It’s not practical, It’s not economical, you probably won’t keep it long term. Build it anyway. You’ll learn like crazy from it. You’ll play with it happily for hours. And in the end, you know you want to.
Besides, you wouldn’t believe some of the stupid stuff I’ve built, just because I can.
I’d also bet that for Amal, Satur9, Pilgrimaster, Devilclarke, and a dozen other people on here too.
We have load cells in the storage area where I work. It’s easier to keep track of loose items likes screws and nuts. Lemme tell you: those things are exceedingly expensive. We have 2 rows of shelving with maybe 20 scales in total, and the sampling station, and that setup is 5 figures.
But then I suppose if you DIY it it would be cheaper. The stuff we have is an industrial solution.
not sure what counts as cheep but I was happy with that. I’ll find the product page later.
and enrolling it into your inventory.