The overall goal of Knowledge4Retail is the development of an open-source platform (K4R platform) for complex AI- and robotics applications in retail. At its core, the platform provides so-called semantic digital twins, which represent a digital copy of a retail store. Semantic digital twins allow the construction of realistic digital worlds, which can be annotated with machine readable background knowledge and linked to other information systems. In this way they provide a powerful basis for AI- and robotics applications in retail. The potential of the K4R platform will be demonstrated in four representative application scenarios.

Duration: 01.01.2020 till 31.12.2022
Donee: German Research Center for Artificial Intelligence GmbH
Sponsor: Federal Ministry for Economic Affairs and Climate Action
German Aerospace Center e.V.
Grant number: 01MK20001B
Partner: team neusta GmbH,
Allgeier Consulting Services GmbH,
neusta GmbH,
Universität Bremen,
fortiss GmbH,
neusta software development West GmbH,
dm-drogerie markt GmbH + Co. KG,
dmTECH GmbH,
Technische Universität München,
Kaptura GmbH & Co KG,
Fraunhofer-Gesellschaft zur Förderung der angewandten Forschung e.V.,
Ubimax GmbH,
EHI Retail Institute GmbH
Application Field: Logistics, Production and Consumer
Related Projects: Hybr‐iT
Hybrid and intelligent human-robot collaboration – Hybrid teams in versatile cyber-physical production environments (11.2016- 10.2019)
Intelligent Human-Robot Collaboration (03.2015- 06.2016)
Hi-Digit Pro 4.0
Flexible High-Lift Wing Assembly with a Digital and Efficient High-Rate Production in the Industry 4.0 (01.2018- 03.2021)
Related Robots: MARLIN
Mobile Autonomous Robot for IntraLogistics in Retail
Related Software: Rock
Robot Construction Kit
Machina Arte Robotum Simulans

Project details

Service robotics for the support of store employees: By exchanging knowledge with the digital twin, intralogistics of the store environment can be optimized and employees will be supported within ergonomic aspects (Photo: IAI, Institute for Artificial Intelligence, University of Bremen)

In order to be competitive with respect to international online-sellers, retailers have to combine competence and customer confidence with the possibilities of digitalization.  In a future retail store, employees are providing advice to customers, while robotic systems are in charge of stock-taking, replenishing and collecting scattered items. Smartphone apps direct the customer to the desired goods and answer their queries related to the assortment. Finally, the technology supports visually impaired or disabled people with shopping.

In order to make full use of the possibilities of digitalization in retail, detailed and comprehensive models of the stores, the selling processes and operation sequences are required. Within Knowlegde4Retail, these are provided via an open-source platform (K4R platform), which serves as basis for complex AI- and robotics applications in a retail branch.  As a core technology, sematic digital twins will be developed within the project.  Sematic digital twins are a generic, machine-readable format for digital representation of retail stores. They allow the construction of realistic digital worlds and enable a variety of novel AI- and robotics applications in retail.

The potential of the K4R platform will be illustrated in four representative applications scenarios:

  • Intelligent intralogistics for retail stores
  • Strategic trade marketing for setting up highly customized stores
  • Service robotics for the support of store employees
  • Internet of Things (IoT) – Connecting an intelligent fridge

Within Knowledge4Retail, the DFKI develops the application “Service robotics for the support of store employees” as an example for usage of the K4R knowledge processing services. For this purpose, a mobile service robot for the support of retail intralogistics will be developed. This robot can be seamlessly integrated into the K4R platform and exchange knowledge. Supportive activities include autonomous transport of goods, stock-check, as well as perception of the store environment. In doing so, the robot extends and improves its capabilities with the help of background knowledge provided by the K4R platform. On the one hand, by knowing locations of goods or shelves, availability of employees or other constraints (e.g. urgency, order of unloading, …) actions and routes can be optimally planned and adapted in case of changes in the environment. On the other hand the system will be able to update the store representation in the digital twin by perceiving the environment through its integrated sensors and comparing it to the current digital representation. This bilateral exchange of knowledge with the digital twin helps making intralogistics more efficient and provides ergonomic benefits to the employees.


Knowledge4Retail: Presorting with the Digital Twin

What will the #retail of the future look like? In Knowledge4Retail (K4R), #DFKI and the project consortium led by team neusta GmbH are developing innovative #ai solutions for stationary retail. The core of the project is the K4R software platform, which creates a link between the digital and analog worlds. The platform enables novel AI applications using a digital representation of the retail store, the so-called semantic digital twin. The video shows one application of the K4R platform, intelligent presorting. Using information from the digital twin, the algorithm developed in the project can calculate the number of goods to be re-stocked based on shelf capacities, forecast sales, and optimize the stocking sequence. This avoids unnecessary goods movements and also ensures that the right products are replenished on the exact day before shelf gaps occur. In addition to shelf replenishment, the technology can also be used when, for example, employees pick goods for collection as part of click&collect. The optimization can reduce walking distances and relieve employees, leaving them more time for customer service. The application shows how artificial intelligence can help make stationary retail more attractive in order to remain competitive with online retail.

Knowledge4Retail: Artificial intelligence for the retail sector

The Knowledge4Retail project shows the opportunities that lie in linking personal advice and customer trust with the possibilities of digitalization for brick-and-mortar retail. Under the leadership of team neusta GmbH, partners from the fields of IT, retail and research - including the Robotics Innovation Center and the Innovative Retail Laboratory at DFKI - have joined forces.



MARLIN: A cloud integrated robotic solution to support intralogistics in retail
Dennis Mronga, Andreas Bresser, Fabian Maas genannt Bermpohl, Adrian Danzglock, Simon Stelter, Alina Hawkin, Hoang Giang Nguyen, Michael Beetz, Frank Kirchner
In Robotics and Autonomous Systems, ScienceDirect, volume o.A., pages o.A., Feb/2024.

Back to the list of projects
last updated 04.01.2024
to top