Artificial intelligence at the fruit and vegetable counter – SPAR is testing a new shopping experience

The aim of the innovative solution is to make the shopping process faster, simpler and more accurate, while modern technologies are integrated into the everyday shopping experience. The launch of the pilot project was based on international experience: the technology has already been tested in Austria by the store chain. The idea of ​​expanding technology is closely aligned with the retail trend that brings increased exposure to automation, digitization and AI-based solutions. According to SPAR’s experience, customers are also increasingly demanding developments that not only make processes easier, but also make shopping faster and even more fun.

The scales installed in the test store are manufactured by Bizerba, and the software running on them is specifically tailored to the needs of the Hungarian market. The system uses offline object recognition based on artificial intelligence: the camera only takes a picture when the customer places a product on the scale. Recognition is done using deep learning image classification algorithms that identify vegetables and fruits based on shape, color and size.

The operation of the solution can also be easily understood by customers. In an ideal case – for example, when weighing a banana – the system clearly recognizes the product, so only one selectable option appears on the display. In less clear situations, such as several types of apples, the scale first identifies the product category and then narrows down the options to the possible varieties. In this case, the customer only has to choose exactly which product he placed on the scale. This logic both speeds up the process and reduces the chance of making mistakes.

“This development is a significant step in the life of SPAR Hungary, but at the same time it is also an important stage in a longer innovation process. Our goal is to gradually integrate modern technologies into our daily operations with the customer experience in mind”

– said Márk Maczelka, head of communications at SPAR Hungary. He added: the pilot project provides an opportunity to refine the solution in a real environment, based on customer feedback, before introducing it on a wider scale. The digital system works in a self-learning manner, i.e. it continuously improves the recognition of new products during use, based on customer feedback and manual clarifications.

The pilot started roughly three months ago, and is currently taking place in a store in Tatabánya. The employees received targeted training, and the customers are helped by the information materials placed next to the scales. The evaluation of the experiences of the test period is still ongoing. Based on feedback and operational data, SPAR prepares a schedule for the further introduction of the technology broken down by store, which can be the next step in a comprehensive digitization development process.