Project overview

Factory planning projects require an abundance of information in a wide variety of formats and
variance is required for the individual elements of the planning object. On the one hand, this is available in
systems, forms of presentation and software applications or can only be obtained through discussions. Sorting and evaluating heterogeneous data volumes with regard to suitable quality and completeness takes a lot of time. With the help of AI approaches (clustering and classification methods), this effort can be significantly reduced compared to the current manual or partially automated approach.
As part of the GenAI4FFD project, an assistance system based on Generative Artificial Intelligence
(GenAI), an assistance system is being developed that provides support in the early phases of the factory planning process
from the beginning of target definition to the completion of detailed planning. In the first step, a factory planning-specific GenAI-based solution will be developed, which will be used as a prototype for three central task areas of factory planning as part of the project:

1. Requirements recording and analysis,
2. Generation of factory designs,
3. Model creation and evaluation
.

Existing European GenAI models are compared, evaluated and specialized for factory planning knowledge. For this purpose, specialist literature, relevant standards and guidelines and, above all, data from completed factory planning projects with operational information on their performance and framework conditions are used. This forms the basis for the three fields of application, which are then implemented prototypically, firstly via the digital room book and secondly as an independent web instance and aggregator in planning and simulation software. At the same time, it will be examined how the solution can be transferred to other planning issues, such as housing development, and how it can be easily used for SMEs.
This text was translated with DeepL on 18/12/2025

View project in the research portal

AI Engineering (AiEng) encompasses the systematic design, development, integration and operation of solutions based on artificial intelligence (AI) using engineering methods as a model. At the same time, AiEng builds a bridge between basic research on AI methods and the engineering sciences and makes the use of AI systematically accessible and available there. The project focuses on the nationwide development of an "AI Engineering" Bachelor's degree program, which combines the training of AI methods, models and technologies with those of engineering sciences. AiEng is to be designed as a cooperative study program between Otto von Guericke University (OVGU) Magdeburg and the four universities in Saxony-Anhalt: Anhalt University of Applied Sciences, Harz University of Applied Sciences, Magdeburg-Stendal University of Applied Sciences and Merseburg University of Applied Sciences. The interdisciplinary degree program will enable students to develop AI systems and services in the industrial environment and beyond and to provide holistic support for the associated engineering process - from problem analysis to commissioning and maintenance / servicing. The AiEng curriculum provides comprehensive AI training, supplemented by basic engineering training and in-depth training in a selected application domain. In order to achieve a symbiosis of AI and engineering education, a new action-oriented framework is developed and taught, which describes the complete engineering process of AI solutions and methodically supports all phases. AIEng is characterized by a cross-module interlocking of teaching and learning content within a semester as well as by a cross-faculty and cross-university tandem teaching concept and pursues a student-centered didactic concept, which is supported by many practice-oriented (team) projects and a wide range of Open Educational Resources (OERs) with an (e)-tutor program.

View project in the research portal

This dissertation explores the application of multi-agent reinforcement learning (MARL) for
handling deadlocks in intralogistics systems that rely on autonomous mobile robots (AMRs).
AMRs enhance operational flexibility but also increase the risk of deadlocks, which degrade system
throughput and reliability. Existing approaches often neglect deadlock handling in the planning
phase and rely on rigid control rules that cannot adapt to dynamic operational conditions.
To address these shortcomings, this work develops a structured methodology for integrating
MARL into logistics planning and operational control. It introduces reference model that explicitly
consider deadlock-capable multi-agent pathfinding (MAPF) problems, enabling systematic
evaluation of MARL strategies. Using grid-based environments and an external simulation
softare, the study compares traditional deadlock handling strategies with MARL-based solutions,
focusing on PPO and IMPALA algorithms under different training and execution modes.
Findings reveal that MARL-based strategies, particularly when combined with centralized
training and decentralized execution (CTDE), outperform rule-based methods in complex,
congested environments. In simpler environments or those with ample spatial freedom, rulebased
methods remain competitive due to their lower computational demands. These results
highlight that MARL provides a flexible and scalable solution for deadlock handling in dynamic
intralogistics scenarios, but requires careful tailoring to the operational context.

View project in the research portal

Within the R&D project "GEOSTALOVR", the development of a simulation-based geodesign platform for planning logistics and tourism in European cities is planned. By means of VR visualization and interaction possibilities, the planning process is to be digitalized throughout, democratized and made more human-centered. The development is being realized in an international cooperation project with Turkey in collaboration with three SME partners (LS Software & Engineering GmbH, Globetech Cografi Bilgi Teknolojileri Yazilim Planlama Harita Enerji Egitim Mühendislik Danismanlik Limited and Elf Sehir Planlama) and two research partners (Otto von Guericke University Magdeburg and Harran University). The project is designed to run for two years. A prototypical setup and testing of the system is planned based on the city of Sanliurfa in Turkey. The project emerged from the network "NekoS" and is accompanied by the network management (ZPVP GmbH) during the implementation.

View project in the research portal

Development, implementation and evaluation of a simulation and test platform for the development and evaluation of algorithms and strategies for innovative warehouse concepts in manufacturing companies.

View project in the research portal

The project aims to enable the LogCentre of the German-Kazakh University (DKU) to provide trainings in the areas of

• Transport
• Business games in production and logistics
• Simulation, optimization and AI for decision support in production and logistics
• Urban and sustainable logistics

on the Kazakh market and to be able to carry them out on a self-supporting basis. The target group of the trainings are companies operating in Kazakhstan and Central Asia as well as governmental organizations such as city administrations and universities.

For this purpose, the following work packages are to be processed jointly by OVGU and DKU in the project:

• Project and risk management
• Needs analysis
• Business plan
• Selection and training of DKU LogCentre personnel
• Creation of concept and materials for the training measures
• Pilot trainings with German-Kazakh trainer tandems in the above mentioned areas
• Commercial trainings in the above mentioned areas
• Communication and marketing

View project in the research portal

ProdLog addresses the issue of a weak industrial sector in Kazakhstan, Kyrgyzstan and Russian Federation and focuses on enabling universities to gain and provide a profound and holistic knowledge on planning and operating sustainable production processes. For that purpose a bologna-based master curriculum with 18 modules in resource efficient production logistics will be developed and implemented in six universities of the partner countries. The academic staff will be trained with innovative teaching methods in the learning factory "Technology centre for production and logistics systems PULS" and  equipped with state of the art logistics laboratories. By means of that, the understanding of logistics shall be widened - away from transport logistics to a systemic and interdisciplinary approach of applicant-oriented education, challenges with economical, political and social problems of our society.

View project in the research portal

Development of material flow simulation models based on Plant Simulation and the development of a connection of these with a digital twin of production plants and a visualization platform with the work packages:
- Basic development of a connection of a material flow simulation with a digital twin of a production plant and basic visualization in a 360 degree visualization system
- Development of an extended material flow simulation

View project in the research portal

The R&D project "LOCSys" (Laundry Order Consolidation System) aims to develop and test an innovative, automatic system for buffering, consolidating and picking customer-related, small-scale laundry orders in industrial laundries.
The development is being realized in a cooperative project involving three SME partners (EBF Dresden GmbH, FRAIMTEC Automation & Anlagenmontage GmbH and Ematik GmbH) and two research partners (Otto von Guericke University, Fraunhofer IFF). The planned project is designed to run for 2 years. A prototype assembly and testing of the system is planned at the Puschendorf Textilservice GmbH site in Flechtigen or Schönebeck.
The proposed project is an R&D project that emerged from the ZIM network "NekoS" and will be supported in its implementation by the network management facility, ZPVP Zentrum für Produkt-, Verfahrens- und Prozessinnovation GmbH - Experimentelle Fabrik Magdeburg.

The "LOCSys" project is funded as a joint project by the BMWi as part of the Central Innovation Program for SMEs (ZIM).
This text was translated with DeepL

View project in the research portal