Menu & Search

How Autonomous Stores Work (AI Solves the Problem of Queuing)

How Autonomous Stores Work (AI Solves the Problem of Queuing)

Queues, no one likes them. Whether you just want a sandwich or have a trolley full of groceries, your main objective is to get in, get served and get out quick.

Retailers are currently testing out some new ideas around cashier-less shopping and self-checkouts. The dream of an autonomous store may become a reality someday sooner than you think.

The biggest shakeup in the last decade has been the advent of online grocery shopping, though it has its flaws. Convenient as it may be, substituted items and best before dates that don’t last long can put some people off. Plus, people still like to view and select their food in person.

So how do we solve the age-old queuing problem at shops? Is artificial intelligence (AI) the answer? Certainly, technology has changed the way people shop, but can it do more?

Amazon Go – The AI Powered Store

Image: Amazon

The Amazon Go store is revolutionary, it’s a groundbreaking showcase that tells the world, this technology exists and it can be done. The idea is simple, you walk in, scan an app on your phone, pick up your shopping items and ‘Just Walk Out’. The store is located in Seattle and people have been flocking there to test it out.

But how does Amazon Go work? That’s the billion dollar question.

According to Amazon:

Our checkout-free shopping experience is made possible by the same types of technologies used in self-driving cars: computer vision, sensor fusion, and deep learning. Our Just Walk Out Technology automatically detects when products are taken from or returned to the shelves and keeps track of them in a virtual cart. When you’re done shopping, you can just leave the store. A little later, we’ll send you a receipt and charge your Amazon account.

Let’s break this down in simple terms.

The ceiling of the store is covered in cameras and is constantly recording. There’s a big supercomputer somewhere running deep learning algorithms (complicated mathematical programs) which teach the big supercomputer to see like a human (computer vision).

  • As you walk in the door and scan your app the technology senses and tracks your every movement.
  • The technology knows what a bottle of water or loaf of bread looks like.
  • It understands that a human has just lifted it from the shelf to a bag.
  • It understands that a human has returned it back on the shelf.
  • When you are done, the information about the items collected are added to your bill and payment is taken from your card automatically.
  • You leave or ‘Just Walk Out’.

In order for all this happen smoothly the software has to work perfectly. Deep learning has advanced in the last few years because of datasets like ImageNet that were collected. A computer has to view thousands of images or videos in order to learn the difference between actions and objects. So in theory, when more people use the Amazon Store, the more machine learning can be done, making the whole system even smarter.

Standard Cognition

Amazon aren’t the only boys on the block. Fellow rivals Standard Cognition are a new startup developing software for cashier-less stores. There are plans in place to try their system in stores across Japan in 2020, with trials beginning in 2019. How exciting!


Another company working on the autonomous checkouts are Aipoly. They’ve been making headlines with their Aipoly Vision app, that helps people who are visually impaired recognize objects and colors. The app uses artificial intelligence to identify objects when you point your phone at something. Another Aipoly app is being developed for autonomous markets using their visual search engine technology for products.

Moby Mart

And there’s more. Moby Mart is an AI powered mobile supermarket, with absolutely no staff. You scan what you want with an app, and are charged as you leave. It’s currently being beta tested in Shanghai. Developed by the Swedish startup Wheelys in the Y Combinator, the prototype is just a glimpse of the future of shopping.

What about the new Scan and Go technology?

Scan and Go is a new concept still in its infancy. Shoppers pick up a special handheld scanner or use an app on their phone. You simply scan each item you want to purchase and can pack directly into a shopping trolley of empty bags. When you’re finished you pay through the app or the Scan as you Shop payment area. The advantage here is that you can see how much you’re spending as you go along and when you’ve finished, your groceries are already packed.

IMAGE: Tesco

There are several big brands already trying this Scan and Go technology out such as Tesco and Sainsburys in the UK. The main problem here is technology adoption. It can take a while for change to be accepted, especially when some people are not as tech savvy as others. Walmart tried Scan and Go, but found customers preferred not to use it, saying there was ‘low participation’. But every town is different, people are different, there is no one size fits all when it comes to grocery shopping. Scan and Go is a trend you might see more of depending on where you live. Walmart is stilling trying shelf-scanning autonomous robots, to help workers stocking shelves.


What does the future hold?

AI is here and the technology behind it is developing at a phenomenal rate. Supermarkets and stores will be doing a lot of testing before such a system is fully implemented. Within the next five years though, you will start to see more AI stores popping up. Customer adoption and cost of integration will play an important role in the decision-making process of brands. Maybe people will prefer Scan and Go, maybe they just want to walk out or maybe they want to queue.