Technical article

Preventing revenue losses: fraud prevention at the self-service scale

A wide variety of technologies based on innovative process automation and digitally supported shopping experiences are booming in the stationary retail sector. In particular, networked retail solutions featuring artificial intelligence (AI) are growing in popularity. AI-based self-service scales with object recognition in the self-service are of the convenience store are just one example.
  • Experience

The 6 benefits of networked self-service scales with object recognition

Reduced revenue shortfalls by minimizing incorrect registrations 

Reduced losses, because the ability to recognize packaging means there is no need for manual selection and use of a standard tare 

Shorter wait times, because the checkout process is sped up 

Eases the load on cashiers because there is no need for weighing and item searching 

Positive customer experience with self-service weighing without having to manually enter information 

Improved sustainability because of unpackaged products, with no added effort 

Produce departments within the retail sector have visibly changed over the last 20 years. Product ranges have increased dramatically in size and the percentage of unpackaged products requiring weighing has increased. Here are some of the reasons:
1
Saving on packaging

Instead of offering products in plastic bags or laminated shells, producers and retailers are increasingly offering unpackaged items for increased sustainability, which are then weighed in recyclable bags used for fruit and vegetables.

2
Globalization

Whether strawberries from Israel or asparagus from Peru – the range of fresh products now available all year round has expanded exponentially.

3
Bio boom

Many different types of produce are displayed twice: conventionally grown and also with a bio label.

4
Regional products

These days, retailers have expanded the standard product range by adding more regional products, freshly harvested and with short delivery paths.

5
Seasonal produce

Rapini, salsify, wild garlic, pumpkin and so on – seasonal classics are back, expanding the product range all year round.

Special offer or Class I goods?

Though this kind of variety is very much welcome on the plate and in the fruit basket, it does make it more difficult and time-consuming for shoppers to select the correct products on the scales. The disadvantages of typical self-service weighing: Given the large variety of items, having to search for the symbols representing the products or enter article numbers that you may have forgotten by the time you arrive at the scale is quite awkward. Add to that, consumers sometime deliberately declare incorrect items, for example those bananas that are on sale instead of the more expensive Class I product.

Initial boom in self-service scales is in the past

Initial results: Following the early popularity of self-service scales with integrated label printer, many supermarkets have gotten rid of them. Instead, products are once again weighed at the counter – a process that takes a lot of time and requires more staff if customers are not to be left waiting. So, what can be done to keep cashiers from having to weigh products, while making the customer experience more convenient, and reducing revenue losses from incorrect input at the same time?

AI does it better

Self-service scales with visual object recognition, supported by AI are the answer. A typically installed camera system records the products placed on the scale. An algorithm analyzes the image data and the result appears on the display in less than a second – shoppers only have to add the item and register it.

Self-learning object recognition system

Object recognition as such occurs entirely offline, directly on the scales. The recorded images are used for evaluation only and are transferred to the cloud as a foundation for continuous improvements to the model. As a result, the system learns from weighing processes in many different supermarkets every single day and provides increasingly reliable results.

Conclusion

An AI-based scale with object recognition is a faster, more reliable solution for declaring items than input by shoppers or cashiers. It is time for the self-service scale to make a comeback in the produce section – for example, the new K3 from Bizerba with object recognition.

Any questions?
Contact us