You've probably seen the Amazon Go promotional video everywhere. The store without cash registers. You go in, scan a QR Code displayed on your phone by the Amazon Go app, take whatever you want from the shelf and leave. Payment is made automatically after you leave the store, through the app.
How it works: Amazon says it has integrated into the solution computer vision, deep learning algorithms (artificial intelligence) and many sensors, located on the shelves. Most likely these are RFID sensors/tags, similar to the anti-theft tags on clothes. To which are added video systems similar to those of Kinect, which "read" gestures, plus an AI infrastructure that interprets all the information.
The concept is not new. Here's a little more advertisement black, published by IBM ten years ago. What was missing from there: the way payment for purchased products is associated with the buyer. That is, the smartphone and the app.
How much does it cost?. The first thing you think about, from the integrator camp, is the costs: sensors, data infrastructure, it seems a bit expensive for a regular store. So a high volume of purchases is important for the investment to make sense. But the Amazon Go solution itself can be a factor in increasing sales volume. It is worth investing in a solution that, even if expensive, brings many more buyers into your store – who enter and exit just as easily.
Advantages you throughout this system: the software solution learns very quickly the behavior of shoppers at the shelf. Amazon GO will be able to quickly offer suggestions for positioning, rearranging merchandise and traffic, exposing to more points of interest to stimulate purchase – things that already happen but somewhat "manually".
Security. What happens if someone else takes your phone into the store? There will most likely be video recordings and associations of these images with the products taken from the shelf. Or a PIN code that you enter into the smartphone app to activate purchases, before leaving the store. There are solutions.
