Amazon Go stores could watch, listen, and remember your every move
There’s a lot of excitement about Amazon Go, the company’s new kind of grocery store that lets customers take what they need and leave without ever waiting in line, but there weren’t a whole lot of specifics on how this system works. An Amazon patent filed in 2014 gives us a glimpse of how this could all operate, and it involves — you betcha — hella cameras.
The patent describes a system where cameras could capture
you as you walk into the store, then identify who you are based on an ID card
that’s associated with your Amazon account (in Amazon Go’s case, an app that
you scan when you enter). Facial recognition may also be used.
Once you’re in the store, cameras could follow you around
as you navigate the aisles to see where you’re browsing and what items you pick
up. “The user's location within the materials handling facility and
identification of the inventory location within the user's location may also be
determined,” the patent details.
“Microphones may record sounds made by the
user and the computing resource(s) may process those sounds to determine a
location of the user.” In addition to microphones, Amazon also listed cellular
triangulation as a potential method for locating a customer’s location.
As you’re shopping, sensors in the inventory shelves
monitor when which items are being picked up. The sensors watch for whether the
item gets put back down on the shelf or not, and may even check if the weight
of the item changes from its original state. This is likely designed so people
are correctly charged if they pick up a food or drink and begin consuming them
while they’re still in the store.
Static cameras watching the shelf also determine whether
multiples of the same item are taken, such as several sticks of gum or a few
bags of chips. If you pick up several bags of chips just to get the one on the
inside, the camera is supposed to know that you ended up putting the other bags
of chips back as well. If Amazon is not confident about how many items you took
or what exactly you took (i.e., a Colgate versus Crest toothbrush) it may end
up confirming with you to charge the right amount.
The cameras can even go so far as to determine your skin
tone color. The patent says this is used to identify the shopper’s hand to see
whether they actually pick up anything off a shelf, but combine that with the
fact that Amazon knows what you’re buying and who you are down to your skin
color and this is pretty next-level market research data. (Not to mention...
kind of creepy?)
The patent provides a good look at how Amazon Go could
work, but there are still some unanswered questions: what if you go shopping
with a friend and they pick up something and hand it to you? Whose account gets
charged for that item? How does Amazon determine if two people who enter the
store together are family members in the same household account? What happens
if they separate in the store and leave at different times? We’ve reached out
to Amazon for clarification, but the company is not currently offering press
interviews.
It’s important to note that these are all just ways
Amazon could implement the technology — this is not the final version of what
is going to be done in-store, which has only opened in beta to Amazon
employees. Still, the patent does describe basically everything the company
teased with its announcement yesterday, so this seems rather close to how
Amazon Go will work in practice. If you’re still curious, feel free to read
through the full patent here.
Keyword: Amazon Go, amazon go store, amazon go locations,
amazon go youtube, amazon go video, how does amazon go work, what is amazon go,
amazon go seattle, 2131 7th ave, seattle, wa, amazon go grocery store, amazon
go location, amazon go technology, amazon go reddit, amazon go grocery, amazon
go store locations, amazon go stores, amazon grab and go, how amazon go works,
amazon go rfid, amazon go store location, amazon go store seattle, amazon go
jobs, amazon go how does it work, amazon gi, amazon go shopping, amazon go nyc
0 Komentar