Just when you thought what the next big leap in retail technology might be, retail giants Amazon has come up with a concept that can potentially rattle the entire retailing landscape. Amazon recently unveiled its' first (of hopefully many) high-tech retail location called 'Amazon Go' in Seattle. Currently in a private testing phase, Amazon plans to make the 1800 square foot convenience grocery store available to the public by early 2017. The primary aim being to speed up the overall buying process by enabling no check-out lines.
Amazon Go – The Next Big Thing
Amazon Go is already being considered as one of the most advanced shopping technology ever created. Named as 'Just walk out' technology, this can effectively enable customers to step into the store, pick up what they want and leave immediately without having to wait at the checkout lines or kiosks. Customers can avail this facility through the Amazon Go app on their supported smart phones. All they have to do is scan the bar code at the entrance and walk in. Whatever they pick up for purchase will automatically be added to the cart. Once the customers are done shopping, they may proceed to the exit point. The exit is auto-detected and all the items in the cart are added for billing. As Amazon already has the customers' credit card details on file, on exit, the bill amount for the items purchased is charged and the receipt generated is directly sent to the app.
The Bundle of Technology involved
It's not the current most talked about breakthrough in retail for nothing. A bundle of technologies like computer vision, artificial intelligence, deep learning algorithms, sensor fusion, camera tracking, etc work in tandem to bring forth the most advanced shopping experience.
Customers can be identified during entry with the help of the bar code scanning through the Amazon Go app. The cameras capture the customers before the app identification is carried out.
After entry, the cameras follow the customers and capture them as they navigate the different aisles to pick up what they want. As they shop, the sensors placed in the inventory shelves monitor the items that are being picked up and also kept back.
For the inventory management system to accurately detect the items that are being picked up, the sensor fusion technology is used. Detecting matching items can be tricky, but data from various input devices helps in determining the exact identity of the item that is picked from or placed back in the inventory locations.
Image Analysis can potentially narrow down the list of matching items, but if assisted with weight analysis, errors in item detection can be eliminated.
The weight of a particular item can be determined with the help of data from a load cell or a pressure sensor placed in the inventory location. The weight of an item can be compared to the stored weight of each matching item for the accurate identification of the item placed in the inventory. Apart from this, the purchase history of the customer can be considered to get an idea of the most likely item that the customers might pick up.
This is how data generated from multiple inputs can help in accurately identifying the item that is being picked up/placed back on the inventory. Once the item picked up for purchase is detected, it is automatically added to the cart in the app.
Static cameras also detect whether multiple numbers of the same item has been picked up. The camera tracking technology apart from facial recognition can also go to the extent of detecting the skin tone color of a customer's hand to determine whether he picked up anything from the shelf.
Amazon Go certainly looks promising. Currently in testing phase, one might imagine that if it even nearly delivers on the fancy promise made, the face and landscape of retail may well be redefined once again.