Retail’s Big Show, the yearly event that shows off the newest technologies in the retail branch, recently closed its doors in New York. Like in many branches, artificial intelligence plays an undeniable role when it comes to digital transformation of retail. Benjamin Thomas, Innovation Consultant, takes a look at three startups who reinvent shopping with the help of AI. Thanks to AI, distributors can completely change their way of working, from offering an improved customer experience to a new way of managing stock.
Shopping carts were used for the first time in the sixties and not much changed since then. Focal Systems, a startup, founded in 2015 by some Stanford graduates, wants to use AI (and more specifically image recognition) to give the carts some superpowers. The first step consists of adding a lateral camera that continuously films shopping shelves while customers are shopping. These images are analyzed, and algorithms detect if the shelves are sufficiently filled. If this isn’t the case, a message is sent to the staff to take action.
The startup takes it even further. With the newest concept, a camera that films the content, is added to the cart. Every time a customer drops something in the cart, the image is analyzed to determine the product. At the end, the customer only has to pay, nothing has to be scanned anymore. So, scans at the check-out counters aren’t necessary anymore. Amazon recently opened its first store without counters, but unlike Amazon’s concept, this one is quite cheap: a camera system is not expensive anymore thanks to the rise of smartphones. Moreover, you don’t have to reorganize the whole store.
Efficient stock management has always been a problem for distributors. How can you make sure that you always have enough items available to respond to the rising demand, without the risk to buy too much items that will not get sold? Since a few years, several instances investigate statistic models to minimize the huge costs that come with stock management. Vekia, a French startup, wants to help retailers with this optimization. The company offers a software solution that makes predictions and that is much more precise than other traditional methods.
The advantage of Vekia? A team of real specialists regarding AI (this is not always the case for many startups that claim being experts in AI / Deep Learning / Machine Learning). The founder is Manuel Davy, a Cambridge University graduate. He worked at the CNRS (Centre National de la Recherche Scientifique) before he founded Vekia. Thanks to his expertise, his startup can feed the algorithms with data from various sources: the weather, social media, buying history, travel time to the store, etc. This increases significantly the precision of the indicators, so distributors can better manage demand changes. The startup, that already employs nearly 60 people, was present at the exhibition to increase its international notoriety.
If there is a domain everyone thinks about nothing can be improved anymore, it is barcode scanning. Barcodes are already used for many years at supermarket check-out counters through the famous barcode scanners, in stock management of large depots or to make package distribution a lot easier. But if you already tried to scan a QR code or a barcode, you will have noticed that it is not a very intuitive process and, in some cases, time-consuming. Scandit, founded by a team of researchers from the famous MIT and the ETH in Zürich, proves that there are still a lot of possibilities in this domain and that this technology, developed in the seventies, can get a second life.
By using advanced algorithms with image recognition, they have multiplied the power of the scanners through smartphones’ cameras. Multiple codes can be read at the same time, so an entire shelf can be scanned at once. A customer can thus check at the blink of an eye if products contain a specific ingredient (e.g. gluten). A shop assistant can also check if the prices mentioned are correct and if a product is still available. In a depot, an employee who searches for a specific package, can point his smartphone to the complete shelf and immediately find that package. This will help him to win a significant amount of time when preparing a package for delivery.
This startup doesn’t want to stop here. They are already collaborating with some other companies, like drone manufacturers. The drones will be equipped with a Scandit module, allowing it to make an inventory in an autonomous way in a record time.