Artificial intelligence for the retail sector


In the Knowledge4Retail project – K4R for short – partners from IT, retail and research have joined forces to revolutionize brick-and-mortar retail stores. Their common objective is to build an AI-based data platform that connects the digital world with stationary retail environments and makes use of digital solutions for individual customer service – all as open source.

Project description

In today’s world, the shopping experience increasingly occurs online, a development that works to the disadvantage of the stationary retail trade. To remain competitive, retailers must create a link between the analog and digital worlds. To date, digital innovations have been driven mainly by global players. Small-to-medium enterprises are almost nowhere to be found in this environment. Customer experiences and customer services differ widely in both worlds. The advantages of the stationary retail business – trust on the part of the customer and the consulting know-how of the retail staff – must be combined with digital services. This calls for detailed, comprehensive digital business models, as well as sales and operations processes that have been unavailable to date. This is where K4R comes in.

The overall goal of the consortium is the establishment of the K4R open platform, which will serve as the core for complex AI-based planning applications, as well as for in-store robotic applications. As a powerful data basis, this open source platform creates “semantic digital twins” (semdZ) of retail stores as the foundation for various AI and robotics applications from different providers. The K4R platform thus reduces the setup time and the cost barriers for retail companies when introducing AI solutions.  In addition, the open standards help lower the barriers to entry for SMEs in the IT industry that have specialized in individual AI applications. The consortium partners are currently working on four use cases: intelligent intra-logistics, optimal store setup, service robots and intelligent refrigeration units.

Research contribution

K4R will lead to the establishment of a new generation of information systems for the retail sector and its supply chains that serve as a digital innovation platform and ecosystem. Semantic digital twins – digital models of real businesses that combine different data within a retail store - are the foundation of the platform. The project thus involves developing a data format for the digital representation of the setup and processes of the retail business and also integrating data from the store’s environment, such as geodata. Standardization makes interaction between data suppliers and solution providers possible. Using this ecosystem as a foundation, developers can improve their own AI applications or combine them with applications from other providers to form new AI applications. The platform’s open standards enable easier access to the retail company IT infrastructures.

A variety of data and different types of data are required to successfully scale this ecosystem. Fortiss is providing support to the project consortium in this area. Strategies and recommendations are developed using tool-supported, scenario-based analyses of the ecosystem in order to facilitate a sustainable and robust project success. The development of recommendations for platform operators, as well as partial strategies for individual stakeholders, help to increase the growth of the platform. In order to simultaneously enable service-wide applications, fortiss is conducting research into the semantic annotation of web services and the trustworthy and verifiable exchange of data along interorganizational business processes in the ecosystem. The specification and prototype implementation of a semantically enriched interface enables knowledge-based inquiries of data from linked web services between highly networked organizations and devices.


Federal Ministry of Economy Affairs and Climate Action (BMWK)

Innovation competition “Artificial Intelligence”

Project duration

01.01.2020 - 31.12.2022

Prof. Dr. Daniel Mendez

Your contact

Prof. Dr. Daniel Mendez

+49 89 3603522 168

More information

Project partner