Hi, I'm a senior majoring in Computer Engineering and I'm thinking of ideas for a senior design project.
Right now, I'm interested in creating a smart grocery cart using the Arduino, RFID Reader, and iPad. At the moment, I have imagined that the shopper will be able to add items to the grocery cart and they will be scanned through the reader, sent to a program on the iPad and calculated for PayPal checkout (receipt will be sent to email).
However, I am still trying to figure out how to prevent stealing and how the shopper will be able to scan fruits, veggies, etc. Any links that you could provide?
Surely the inability to accurately detect what has been added to the trolley is a show stopper. Unless you've found a solution to that problem, I don't see how you can present anything credible. This isn't an Arduino problem or an electronic/electrical problem - you need to invent some new sensing technology. If you've done that, you're probably set for life if you can make it work. Without it - do you actually have a project?
To help eliminate stealing you could have electromagnetic system tags incorporated into the tag to be scanned or inside the packaging under the tag.
Fruits and veggies could work if they were pre-packaged. Like carrots, potatoes, etc..
It could also work on weight. Adding an item to the trolley without scanning it would cause an alarm. You'd have to know the weight of every item to ensure it is put in the trolley after scanning and not replaced by a different, more expensive item. The shopper can weigh their fruit, etc. themselves and have the scales print out a sticky barcode label with the weight and cost encoded in it. Weighing 3 apples but putting 4 into the bag before sealing with the label would mean that the barcoded weight would not agree with the weight added to the trolley.
Your problem would be to keep the weighing device in the trolley accurate over a large range.
I'm curious about the RFID aspect of your project. Self-checkout counters at grocery stores do not rely on RFID cards/readers, but they are able to perform the same task you are envisioning, except for the mobility part.
Why not scan the UPC codes, like grocery stores currently do?
What I'd like to see is a running total of what is in the cart, so when I get to the checkout counter, there are no surprises. Countless times, we've gone to checkout at CostCo, and found the bill was more than twice what I expected. When you seem to remember dropping $100 worth of stuff in the cart, and find you've dropped $275 worth of stuff, that's a bit of a rude surprise.
I don't think any stores would be interred in such a product as they would probably lose a lot of money if people have a live up dated price. It would be better if it gave them the total at the end.
Also how about vision tracking? Could you use a camera to see what gets put in the cart? Maybe a weight and camera combo?
Problem with RFID is most items do not have RFID tags, and RFID tags cost money. Barcodes are already on (almost) everything and they're practically free.
I'm not sure putting a scale in each cart is cost effective or accurate -- scales need to be level and motionless to be accurate. What I would do is have a cart scale that weighs the whole cart just before the customer leaves the store (subtracting, obviously, for the weight of the cart). It can compare the net cart weight with the cumulative weights of the items individually scanned previously. Fruits and veggies can be processed on a separate small scale with a touchscreen menu interface (to select the fruit/veggie type) before cart weigh-in.
Or instead of barcodes/RFID, how about an IR sensor in front of each item on the shelf. When an item is detected as being removed from the shelf, the nearest iPad registers the item as purchase pending. (The shopper can return the item to the shelf to clear it back off their list -- needed for folks just reading labels.) Only trick then is deciding which customer picked up the item... hopefully your customers aren't packed in so tight an item removal is registered on the wrong iPad.
tylernt:
Problem with RFID is most items do not have RFID tags, and RFID tags cost money. Barcodes are already on (almost) everything and they're practically free.
I'm not sure putting a scale in each cart is cost effective or accurate -- scales need to be level and motionless to be accurate. What I would do is have a cart scale that weighs the whole cart just before the customer leaves the store (subtracting, obviously, for the weight of the cart). It can compare the net cart weight with the cumulative weights of the items individually scanned previously. Fruits and veggies can be processed on a separate small scale with a touchscreen menu interface (to select the fruit/veggie type) before cart weigh-in.
Or instead of barcodes/RFID, how about an IR sensor in front of each item on the shelf. When an item is detected as being removed from the shelf, the nearest iPad registers the item as purchase pending. (The shopper can return the item to the shelf to clear it back off their list -- needed for folks just reading labels.) Only trick then is deciding which customer picked up the item... hopefully your customers aren't packed in so tight an item removal is registered on the wrong iPad.
Maybe I have the mind of a thief, but your IR idea is easily circumvented. Imagine two bottles of wine next to each other on the shelf. One is an expensive, Chateau grown wine, the other is cheap 'plonk'. It would be easy to slide the expensive bottle sideways along the shelf until it's in front of the sensor for the cheap wine and then remove it. Weighing the bottle wouldn't be much use, as the bottles will probably weigh the same.
Henry_Best:
Maybe I have the mind of a thief, but your IR idea is easily circumvented. Imagine two bottles of wine next to each other on the shelf.
Yeah this is a tough nut to crack. You'd need dividers between items, but even so they can take both bottles and put the cheap bottle back in the expensive slot.
So...would the iPads belong to the store? Or, would the carts have a "dock" for customer to use their own? If the latter, then the app could incorporate a 'shopping list' and 'map' features to also help customer find what they need, and make sure they don't forget anything. Maybe a small generator attached to wheels to help keep everything powered, too?
I like a combination of some of the ideas -
• iPad & iPhones can read barcodes already. Use that.
• To keep it simple, probably just a menu for fruit and stuff (similar to how self checkout works now) w/either small scale on cart, or Bluetooth scales near the fruit stand?
• While shopping (checkout) the iPad app associates to the specific cart ID# (through the dock connector) and keeps track of that for the next step - (and possibly to help connect the fruit scales to the correct ipad?)
• "Truck stop" drive-over style scales that weigh the cart on way out of store. If weight doesn't match what was paid for, wheels lock and attendant is signaled to "help" you. Scale uses RFID in the cart to associate it to the proper transaction from iPad app. (I know here in Chicago, we already have carts that lock the wheels if the cart is taken off the property. So, that tech is already in place.)
tylernt:
Henry_Best:
Maybe I have the mind of a thief, but your IR idea is easily circumvented. Imagine two bottles of wine next to each other on the shelf.
Yeah this is a tough nut to crack. You'd need dividers between items, but even so they can take both bottles and put the cheap bottle back in the expensive slot.
tylernt:
Henry_Best:
Maybe I have the mind of a thief, but your IR idea is easily circumvented. Imagine two bottles of wine next to each other on the shelf.
Yeah this is a tough nut to crack. You'd need dividers between items, but even so they can take both bottles and put the cheap bottle back in the expensive slot.
tylernt:
Henry_Best:
Maybe I have the mind of a thief, but your IR idea is easily circumvented. Imagine two bottles of wine next to each other on the shelf.
Yeah this is a tough nut to crack. You'd need dividers between items, but even so they can take both bottles and put the cheap bottle back in the expensive slot.
