The traditional, stationary trade is increasingly being replaced by online shops. One competitive advantage that mail order companies are using is the possibility of recording the purchasing behaviour of their customers in detail. For example, information about viewed products, the contents of the shopping cart or the shopping history is collected. This data can then be analyzed immediately during the purchase and used, for example, for real-time recommendations of suitable products.
In stationary retail, such data collection has so far only been possible at great expense or only retrospectively, for example using customer cards. A real-time analysis as in online trading is impossible. This is precisely what the VICAR research project aims to make possible.
The aim of the VICAR project is to record the purchasing paths of customers in stationary retail, to analyse them in real time and, based on this, to offer added value to the customers as well as to retailers. From the customer’s point of view, this added value consists of individual recommendations and adaptive personnel planning that offers individual advice exactly when the customer actually needs it. From the retailer’s point of view, customer flows can be visualized, products can be optimally placed, staff can be deployed more efficiently and anomalies such as attempted thefts can be detected automatically.
This is based on video data recorded by video cameras, which are already available for theft protection. The recording of the movement paths and their evaluation is carried out by innovative methods from the fields of machine vision and learning. Special attention is also paid to the anonymisation of the stored data in compliance with data protection regulations. All this should contribute to compensate the current disadvantages of stationary trade compared to online trade and to keep it competitive in the long run.
The VICAR project consortium consists of research, development and application partners. The AWS-Institute, as a partner in the field of research transfer, will contribute the technology for recording the movement paths, which forms the basis for all connected analysis.
The main challenges lie in the real-time processing of the immense amounts of data, the comparison between the numerous video cameras used and the reliable differentiation of large numbers of people.
The VICAR project is funded by the Federal Ministry of Education and Research.