Achtung:

Sie haben Javascript deaktiviert!
Sie haben versucht eine Funktion zu nutzen, die nur mit Javascript möglich ist. Um sämtliche Funktionalitäten unserer Internetseite zu nutzen, aktivieren Sie bitte Javascript in Ihrem Browser.

Info-Icon This content is partly available in English
Show image information
Show image information
Show image information
Show image information
Show image information
Show image information
Show image information
Show image information
Show image information
Show image information
Show image information

Photo: LDM

Photo: LDM

Photo: @AdobeStock/Gorodenkoff

Photo: © AdobeStock/Gorodenkoff

Photo: @ Fraunhofer IOSB-INA

Photo: © AdobeStock/Gorodenkoff

Photo: @ Fraunhofer IEM

Photo: @ Heinz Nixdorf Institut

Photo: @ Heinz Nixdorf Institut

Photo: @ Heinz Nixdorf Institut

Photo: @AdobeStock/Gorodenkoff

Key research area "Intelligent Technical Systems"

Intelligent technical systems are highly complex products, characterized by a close interaction between hard- and software. System functionality is continually improving by moving functionality into software modules. This increases the life span as well as the adaptability of the products, but simultaneously puts high demands on software quality, and especially on the integration with the hardware that is to be controlled.

The underlying research areas that form the foundation for this increased quality demand, are mechatronics, software quality, virtual prototyping/simulation, and systems integration.

Due to a close cooperation between mechanical engineering, electrical engineering, and computer science, we can research novel development methods, in particular in the areas of modelling, simulation, and formal verification. These methods take into account the increasing move from hardware to software, as well as the highly networked system components.

Large collaborative projects

Collaborative Research Centre (CRC) 901 "On-The-Fly Computing"

Individualized IT-services in dynamic markets

Duration: 01.07.2011 - 30.06.2023
Total project volume (University): 30 million euros
Funded by: German Research Foundation (DFG)

The objective of CRC 901 – On-The-Fly Computing (OTF Computing) – is to develop techniques and processes for automatic on-the-fly configuration and provision of individual IT services out of base services that are available on world-wide markets. In addition to the configuration by special OTF service providers and the provision by so-called OTF Compute Centers, this involves developing methods for quality assurance and the protection of participating clients and providers, methods for the target-oriented further development of markets, and methods to support the interaction of the participants in dynamically changing markets.

CRC 901 is divided into four project areas. Project area A is concerned with the algorithmic and economic basics for organizing large dynamic markets. This involves, on the one hand, algorithmic processes for organizing large networks in general and the interaction in networks in particular and, on the other hand, economic concepts for incentive systems in order to direct the participants in the markets. Project area B investigates processes of the modeling, composition and quality analysis of services and service configurations aiming at on-the-fly development of high-quality IT services. Project area C develops reliable execution environments for on-the-fly computing and is concerned with questions of the stability and security of markets, the organization of highly heterogeneous OTF Compute Centers and the provision of configured services by those centers. This project area also involves an application project that deals with systems for optimizing supply and logistic networks, which is considered a long-term application field for the results of the CRC. Project area T bundles the transfer projects of our CRC, which provide a framework for joint research by our CRC researchers and external partners and facilitate the exchange of results from applied research back into our basic research.

Project management: Prof. Dr. Friedhelm Meyer auf der Heide, Heinz Nixdorf Institute of Paderborn University

Projekt partners: various chairs of the Department of Computer Science of the Faculty of Computer Science, Electrical Engineering and Mathematics of the Paderborn University, various chairs of the Departments  1, 3 and 4 of the Faculty of Business Administration and Economics of the Paderborn University, BaER-Lab Business and Economic Research Laboratory, C-LAB Cooperative Computing and Communication Laboratory, DaSCo Paderborn Institute for Data Science and Scientific Computing, IEM Fraunhofer Institute for Mechatronic Systems Design, IFIM Institute for Industrial Mathematics, PC2 Paderborn Center for Parallel Computing, SI-Lab Software Innovation Lab and Weidmüller Interface GmbH & Co. KG and Diebold Nixdorf Systems GmbH

Electronic-Photonic Integrated Systems for Ultrafast Signal Processing (SPP2111)

Duration: 2018 - 2024
Total project volume: 12 million euros
Project volume of the University: 1.522.000 euros
Funded by: German Research Foundation (DFG)

The goal of this priority program is to disrupt the limits of electronic signal processing using photonic-electronic integration in advanced photonic-electronic semiconductor technologies, such as silicon-on-insulator (SOI), silicon nitride (SiN), and indium phosphide (InP). Ultra-fast and energy-efficient information processing is required in many applications, such as communication systems, cloud computing, artificial intelligence, smart factory, instrumentation, and medical technology.

Besides speed and energy efficiency, these systems have unique properties like:

  • Low cost
  • Miniaturization
  • Robustness
  • Programmability

Research on nanophotonic-nanoelectronic circuits and systems will not only increase signal processing speed and enable novel systems, but also significantly improve energy efficiency, thus helping to conserve natural resources and minimize the impact of today's information and communications technology on climate and the environment.

Project management: Prof. Dr. J. Christoph Scheytt, Heinz Nixdorf Institute of Paderborn University

Project partners: Prof. Dr.-Ing. Manfred Berroth (University of Stuttgart), Professor Dr.-Ing. Stephan Pachnicke (Christian-Albrechts University of Kiel), Professor Dr. Jeremy Witzens, Ph.D. (RWTH AACHEN), Professor Dr.-Ing. Christoph Scheytt (Paderborn University), Professor Dr. Thomas Schneider (TU Darmstadt), Professor Dr. Ronald Freund (TU Berlin), Professor Dr.-Ing. Norbert Hanik (TU München), Professor Dr.-Ing. Lars Zimmermann (TU Berlin), Professor Dr.-Ing. Dietmar Kissinger (University of Ulm), Professor Dr.-Ing. Robert Weigel (Friedrich Alexander University of Erlangen-Nürnberg), Professor Dr.-Ing. Frank Ellinger (TU Dresden), Professor Dr.-Ing. Dirk Plettemeier (TU Dresden), Professor Dr.-Ing. Sebastian Randel (KIT), Professor Dr.-Ing. Martin Schell (Fraunhofer Institute for Telecommunications), Professor Dr.-Ing. Christian Koos (KIT), Professor Dr.-Ing. Thomas Zwick (KIT), Professor Dr.-Ing. Franz Xaver Kärtner (University of Hamburg)

AI Marketplace

AI platform for the projects of tomorrow

Duration: 01.01.2020 - 31.12.2022
Total project volume: 16.6 million euros
Project volume of the University: 2.24 million euros
Funded by: Federal Ministry for Economic Affairs and Energy

Artificial intelligence in product creation is an important key to intelligent products and effective manufacturing. With the AI Marketplace, a unique ecosystem is emerging with which companies can unlock the potential in this area. At its heart is a digital platform where suppliers, users and experts can develop and exchange solutions for AI. The vision is a marketplace that offers not only an intelligent matchmaking service but also a protected space for secure data exchange and data sovereignty. This will be complemented by an app store for AI solutions and a range of comprehensive AI building blocks.

The guarantee for success is a project consortium of 20 research institutions, networks and companies, whose nucleus is it's OWL. Other networks ensure a broad impact. Prostep ivip, for example, bundles know-how in product development, and the 'International Data Spaces Association' guarantees secure data spaces. The open source platform 'FIWARE' and the platform operator 'inno-focus' are leaders in their fields of platform development.

The AI marketplace makes an important contribution to making AI solutions available to SMEs. Industrial companies at the OstWestfalenLippe location as well as in all of Germany will become competitive and the global visibility of Germany in the field of artificial intelligence will increase.

Project management: Leon Özcan, Heinz Nixdorf Institute of Paderborn University

Projeckt partners: Center for Cognitive Interaction Technology (CITEC), Claas KGaA mbH, CONTACT Software, Diebold Nixdorf, düspohl Maschinenbau GmbH, FIWARE Foundation, Fraunhofer Institute for Mechatronic Systems Design (IEM), Fraunhofer Institute of Optronics, System Technologies and Image Exploitation, Industrial Automation branch (IOSB-INA), Fraunhofer Institute for Production Systems and Design Technology (IPK), Heinz Nixdorf Institute (HNI), HELLA Aglaia Mobile Vision GmbH, Hella Gutmann Solutions, Institute Industrial IT (inIT), inno-focus businessconsulting gmbh, International Data Spaces Association, it's OWL Clustermanagement GmbH, ProSTEP iViP, Ubermetrics Technologies GmbH, UNITY AG, Westaflex GmbH

Development Platform and Ecosystem for Scalable Special Processors in Edge Computing (Scale4Edge)

Duration: 2020 - 2023
Total project volume: 17.3 million euros
Project volume of Paderborn University: 857,000 euros
Funded by: Federal Ministry of Education and Research

For future tasks such as autonomous driving or Industry 4.0, ever larger amounts of data from an increasing number of sensors must be analysed in the shortest possible time with the help of complex algorithms and artificial intelligence (AI). However, the corresponding processors must not only meet high requirements in terms of computing power, but also in terms of energy efficiency, reliability, robustness and safety, which go far beyond current possibilities. The BMBF's ZuSE projects are designed to meet the user industries' urgent need for future-proof, trustworthy processors that are tailored to their specific tasks and offer high performance.

The Scale4Edge project is researching how the development time and costs of application-specific edge processors can be significantly reduced. Such processors usually perform crucial initial calculations while mobile and close to sensors, at the interface from the real to the virtual world. They must therefore not only be particularly reliable, performant and robust, but also energy-efficient. Furthermore, they must offer a high degree of trustworthiness. With the emerging scalable and flexibly expandable development platform based on the licence-free, open-source RISC-V instruction set architecture, individual processors with these properties can be developed efficiently and cost-effectively.

Project management: apl. Prof. Dr. Wolfgang Müller, Heinz Nixdorf Institute at Paderborn University

Project partner: Infineon Technologies AG, oncept engineering GmbH ASIC- und Softwaretechnologie, TU Kaiserslautern, AbsInt Angewandte Informatik GmbH, Robert Bosch GmbH, Eberhard-Karls-Universität Tübingen, OFFIS e.V., TU München, Albert-Ludwigs-Universität Freiburg, IHP GmbH, MINRES GmbH, TU Dresden, ARQUIMEA Deutschland, SYSGO GmbH, TU Darmstadt, EPOS GmbH, Universität Bremen, FZI Forschungszentrum Informatik

Research Group "Acoustic Sensor Networks"

Duration: 2020 - 2023
Total project volume: 1.6 million euros
Project volume of the University: 1.1 million euros
Funded by: German Research Foundation (DFG)

Distributed Acoustic Signal Processing over Wireless Sensor Networks

The project "Distributed Acoustic Signal Processing over Wireless Sensor Networks" conducts research on signal processing anc communication over acoustic sensor networks. Such an acoustic sensor network consists of nodes with microphones (possibly also loudspeakers) and communicates using wireless transmission techniques. The sensor network is typically connected to the Internet via one or more gateways. Such a network carries out acoustic applications (such as noise reduction and speaker separation). An obvious approach would be to transport all acoustic data collected by the sensor nodes to the gateway for processing. However, this is not necessarily the best possible approach (because of the high communication load, possibly large latency). It also gives away the potential benefit of processing the data already on the sensor nodes, thus reducing data volumes and latency. We are therefore developing approaches (inspired by trends such as microservices and network function virtualization) where acoustic signal processing can be broken into individual processing blocks and distributed over the individual nodes. In this follow-on project, we intend to continue our work on both distributed acoustic signal processing and the development of a framework for automated distribution of such blocks over the network. Specifically, we are looking at hardware aspects (especially with full-duplex audio capabilities) and the possibilities of using such hardware for synchronization in time-based MAC protocols or for estimating acoustic round trip times. This allows us to estimate and appropriately calibrate the spatial geometry of the setup in both static and dynamic environments. Building on information about the acoustic utility of individual nodes, we will address the problem of source selection (signals from which microphones should be included?) both from an  acoustic and a network perspective. A key question will be how to handle such network aspects in dynamic scenarios. Dynamicity arise in particular from movement of acoustic sources and recording devices; such movements can be uncontrolled or controlled (e.g. in the case of robotic sensor nodes). Our work will result in a  testbed that integrates the essential functions and will be an experimental environment for practical testing of the developed algorithms.

Project management: Dr.-Ing. Jörg Schmalenströer, Paderborn University

Coordination project

In daily life, we are surrounded by a multitude of noises and other acoustic events. Nevertheless we are able to effortlessly converse in such an environment, retrieve a desired voice while disregarding others, or draw conclusions about the composition of the environment and activities therein, given the observed sound scene. A technical system with similar capabilities would find numerous applications in fields as diverse as ambient assisted living, personal communications, and surveillance. With the continuously decreasing cost of acoustic sensors and the pervasiveness of wireless networks and mobile devices, the technological infrastructure of wireless acoustic sensor networks is available, and the bottleneck for unleashing new applications is clearly on the algorithmic side.

This Research Unit aims at rendering acoustic signal processing and classification over acoustic sensor networks more 'intelligent', more adaptive to the variability of acoustic environments and sensor configurations, less dependent on supervision, and at the same time more trustworthy for the users. This will pave the way for a new array of applications which combine advanced acoustic signal processing with semantic analysis of audio. The project objectives will be achieved by adopting a three-layer approach treating communication and synchronization aspects on the lower, signal extraction and enhancement on the middle, and acoustic scene classification and interpretation on the upper layer. We aim at applying a consistent methodology for optimization across these layers and a unification of advanced statistical signal processing with Bayesian learning and other machine learning techniques. Thus, the project is dedicated to pioneering work towards a generic and versatile framework for the integration of acoustic sensor networks in several classes of state-of-the-art and emerging applications.

Project management: Prof. Dr. Reinhold Häb-Umbach, Paderborn University

Sound Recognition with Limited Supervision over Sensor Networks

A fundamental problem for many machine learning methods is a discrepancy between the training data and the test data in a later application, which can lead to a significant drop in the classification rate. In acoustic event detection and scene classification in acoustic sensor networks, this problem is exacerbated because there is a very large number of possible sounds and because such networks can be deployed in widely varying geometric configurations and environments. For this reason, existing databases for acoustic event and scene classification will virtually never be a perfect fit for a new application in an acoustic sensor network.

The main goal of this project is therefore to develop methods that allow existing databases to be usable for distinct audio classification tasks in an acoustic sensor network despite this discrepancy. We assume that weakly annotated data, i.e., annotated only with the event class but not with timestamps, are available from another domain, and that unannotated data is available from the target domain. Procedures are now being developed to compute a strong annotation from a weak annotation in which start and end times of acoustic events are additionally annotated to compute domain invariant features, as well as procedures to perform domain adaptation to overcome differences between training data and test scenario in this way. We are also investigating adaptation at test time to adapt to changing acoustic environments and sensor configurations. In particular, deep generative neural models will be used. Appropriate network structures and objective functions will be developed to separate the different factors influencing the observed waveform, in particular the variation caused by the acoustic event from the variation of the signal caused by the environment. Furthermore, we will develop methods to detect unusual acoustic events, because these can be of particular importance for a distinct application. 

Project management: Prof. Dr. Reinhold Häb-Umbach, Paderborn University

Project partners: Dr.-Ing. habil. Gerald Enzner and Prof. Dr.-Ing. Rainer Martin, Ruhr-Universität Bochum (RUB), Prof. Dr.-Ing. Walter Kellermann, Friedrich-Alexander Universität Erlangen-Nürnberg (FAU)

NRW Research College: Light, efficient, mobile (LEM)

Energy- and cost-efficient extreme lightweight design with hybrid materials

Duration: 2014 - 2022
Total project volume (University): 4.9 million euros
Funded by: Ministry of Culture and Science of the State of NRW

The Forschungskolleg LEM is motivated by the use of hybrid components and compound materials: New methods of design, development and production enable the targeted reduction of the masses of components in mechanical engineering, plant and automative engineering. In the context of enhanced recycling and extended reuse potential, the college supports companies by new methods for the selection and introduction of ecologically sustainable resources. The project of the „Product Creation“ workgroup takes over a core role: Requirements from economically driven modularisation of products and resource efficient lightweight design are often contradictory. This contradiction will be addressed by complexity management to find optimal trade-offs between both. One focus is on the design of solutions for lightweight modular design: Methods and digital tools in product development should enable Design for Recycling of lightweight and at the same time modularized mechatronic products.

Project management: Prof. Dr. Thomas Tröster, Institute for Lightweight Design with Hybrid Systems at Paderborn University (ILH) and Prof. Dr.-Ing. Iris Gräßler, Heinz Nixdorf Institute of Paderborn University

Projekt partners: other chairs of the Institute for Lightweight Design with Hybrid Systems at Paderborn University (ILH), Benteler Automobiltechnik GmbH, Boeing, IHK Bielefeld, InnoZent OWL e.V., Kunststoffland NRW, Clustermanager NanoMikro-WerkstoffePhotonik.NRW

NRW Research College Work 4.0: Design of flexible working environments

People-centered use of cyber-physical systems in Industry 4.0

Duration: 2014 - 2022
Total project volume: 4.7 million euros
Project volume of the University: 2,82 million euros
Funded by: Ministry of Culture and Science of the State of NRW

In the NRW Research College “Design of Flexible Working Environments” the effects of Industry 4.0 on the world of work and the role of humans are investigated. The challenge lies in the development of new, social infrastructures. This must anticipate the continuing rapid technological development and see people in the focus of development throughout their entire working lives. To this end, engineers will use methods and tools in the future that will make it possible to consider Work 4.0 as an integral part of product development. Integration into model-based product development is one of the scientific challenges of the Research Training Group. As a specific example, solutions are created for the human-centered, learning-promoting design of assistance systems in the assembly of mechatronic products.

Project management: Prof. Dr. Eckhard Steffen, Paderborn Center for Advanced Studies (PACE) and Prof. Dr.-Ing. Iris Gräßler, Heinz Nixdorf Institute of Paderborn University

Project partners: other chairs of the Paderborn Center for Advanced Studies (PACE), University Bielefeld,  it's OWL, Technologieberatungsstelle DGB NRW e.V., IG Metall NRW, Innovationsnetzwerk Energie Impuls OWL e.V., VDI

AI in the working World of industrial SMEs (Arbeitswelt+)

Duration: November 2020 - October 2025
Total project volume: 10.7 million euros
Project volume of the University: 743,000 euros
Funded by: German Federal Ministry of Education and Research

Artificial Intelligence technology (AI) is associated with far-reaching potential and opportunities for transforming industrial value creation. At the same time, AI is still perceived as a primarily technical option. An understanding of AI in the work context as a comprehensive socio-technical challenge has been rudimentarily established. To date, however, in the AI context, there is a lack of holistic and SMEs close labor research, that provides solution and application knowledge. This discrepancy is being addressed in this project. A basis is a comprehensive approach to the topic by linking People – Organization – Technology and a solution- and transfer-oriented focus in the implementation. The goal is a regional competence center „AI in the working World of industrial SMEs” (in short: Arbeitswelt+). It is intended to serve as a contact point for companies and all the other stakeholders of the industrial work environment.

Competence management, employee participation, and technology acceptance

Artificial Intelligence will fundamentally change the work environment: AI systems support work processes, take over tasks, and create new fields of work. The identification of possible applications and the development of concrete solutions present challenges to small- and medium-sized enterprises, such as lack of professionals or unclear organizational and technological requirements. The Arbeitswelt+ competence center brings together findings from labor research in this future field. The key topics include for example workplace design, competence development, and change management. In pilot projects, research institutions and companies develop concrete solutions in which AI technologies are made available for different fields of application.

Transfer to mid-tier

The results and experiences from the pilot projects should be made available to small- and medium-sized enterprises. To achieve this, an information platform is being developed, best practices are being prepared, and events and workshops are being organized. Employees are qualified for the use of AI technologies in further training courses. In pilot projects, companies can use new AI technologies in cooperation with a research institution to solve specific challenges in their business. Transfer partners of the competence center such as owl maschinenbau and OstWestfalenLippe GmbH provide support.

Project management: Prof. Dr. Kirsten Thommes, Paderborn University

Transregional Collaborative Research Centre TRR 318 "Constructing Explainability"

Explainability of Artificial Intelligence (AI)

Duration: 01.07.2021 - 30.06.2025
Total project volume: 14 million euros
Project volume of the University: 8.2 million euros
Funded by: German Research Foundation (DFG)

How can humans make sense of decisions made by machines? What do algorithmic approaches tell us? How can artificial intelligence (AI) become something that is understandable?

It is frequently the case that technical explanations require prior knowledge about how AI works and are difficult to follow. In the Transregional Collaborative Research Centre “Constructing Explainability” (TRR 318), researchers are exploring how to integrate users in explanatory processes.

The interdisciplinary research team is approaching this topic from two angles: first by understanding the mechanisms, principles, and social practices behind explanations, and second, by considering how this can be designed into artificial intelligence systems. The goal of the project is to make explanatory processes more intelligible and to create easily understandable assistive systems.

A total of 21 project leads, supported by some 30 researchers at Bielefeld University and Paderborn University from a wide range of fields spanning from linguistics, psychology, and media studies to sociology, economics, and computer science are investigating the co-construction of explanations.

Project management: Prof. Dr. Katharina Rohlfing, Paderborn University (Speaker); Prof. Dr. Philipp Cimiano, University Bielefeld (Co-Speaker)

Further projects

Microgrid Laboratory

Duration: 2019 - 2022
Total project volume: 3.7 million euros
Project volume of the University: 3.4 million euros
Funded by: European Regional Development Fund (ERDF)

Today's energy supply system is characterized by grid-connected, geographically distributed structures that must meet the highest safety and reliability standards. The transformation of this system to a sustainable structure characterized by renewable energies is a central societal challenge of the 21st century. The inherent volatility of renewable energy sources requires a move away from hierarchically structured top-down energy grids towards flexible, cross-sectoral and intelligent energy systems by means of a cellular approach. Therefore, in the course of the energy transition, so-called microgrids represent an important solution component to ensure a secure, clean, efficient and cost-effective energy supply in the future. The term microgrid refers to the concept of a local grid consisting of energy sources, storage devices and loads from different sectors, which operates with or without external grid connection. This structure creates a variety of flexibilization options in operation. The local integration of renewable energies by means of microgrids, for example within industrial companies or residential quarters, relieves the distribution and transmission grids and reduces the need for cost-intensive and resource-intensive grid expansion. The efficiency of the energy supply is also increased, since the lossy transport over long distances is avoided and the energy is increasingly generated and consumed locally. Through local storage integration, microgrids can also provide grid-serving services within primary, secondary and tertiary control and even operate autonomously in emergencies as so-called island grids. These grid-stabilizing measures can be enhanced if geographically neighbouring microgrids are coupled to form virtual power plants or large-scale storage facilities. The potential of microgrids has so far been investigated worldwide mainly in academic circles. However, industrial implementation is fraught with high technical and financial risks, especially for SMEs. For successful transfer to industry, however, both extensive practical studies and the upgrading of microgrid components (e.g. power-to-X technologies) for field use are essential. In order for NRW to benefit from the enormous value creation potential of this technology field in a competitive world market in the future, R&D efforts must be intensified and the transfer of knowledge to industry must be strengthened. To this end, the Competence Center for Sustainable Energy Technology (KET) is setting up the "Microgrid Laboratory" under the leadership of the Department of Power Electronics and Electrical Drives. The core of the concept is the development and establishment of a highly flexible, modular development and validation platform for component-related and systemic microgrid research in NRW. High-performance grid nodes are to be developed for the construction of the laboratory. These are freely configurable and flexibly controllable. By means of suitable software, they exactly reproduce the behavior of any components, e.g. batteries, wind turbines or CHPs. The network nodes are interconnected by controllable switches in freely selectable topologies. This hardware and software reconfiguration allows practical research within local grids up to the megawatt range.

Project management: Prof. Dr.-Ing. Joachim Böcker, Paderborn University

Project partners: other KET chairs (EIM-NEK, MB-FVT, MB-TheT) of the Paderborn University

Instrument for pattern-based planning of hybrid value creation and work for the delivery of smart services (IMPRESS)

Duration: 2019 - 2022
Total project volume: 3.2 million euros
Project volume of the University: 782,807 euros
Funded by: Federal Ministry of Education and Research and European Social Fund

The aim of the research project IMPRESS is to develop a set of solution-patterns for pattern-based planning of hybrid value creation and work for the provision of smart services. It should enable companies to shape the transformation independently and purposefully from product manufacturer to smart service provider. For this purpose, methods and tools will be developed which are based on universally valid solution patterns. They show the user proven partial solutions for the design of value creation and work in the context of smart services. The set of solution-pattern, methods and tools will be tested and validated in four pilot projects with industry partners based on their specific smart service applications. The transfer of the partial results will be continuously promoted in workshops with associated SMEs, a transfer group, via multipliers, events and publications. The research project, which consists of ten partners, is funded by the Federal Ministry of Education and Research and the European Social Fund as part of the Future of Work program. The project, which will run for three and a half years, is supervised by the Karlsruhe project management organization. 

Project management: Prof. Dr.-Ing. Roman Dumitrescu, Heinz Nixdorf Institute of Paderborn University

Sociotechnical Risk Management for the Implementation of Industry 4.0 (SORISMA)

Duration: 2019 - 2022
Total project volume: 2.7 million euros
Project volume of the University: 579,285 euros
Funded by: European Regional Development Fund

Industry 4.0 creates significant potential for companies to shape their processes and services along the entire value chain in a more efficient, flexible and resource-saving way. Despite promising opportunities, companies hesitate to use new technologies. The reasons for this are not just technical challenges, but also risks that are difficult to estimate in terms of their organization and employees. After all, beneficial use depends to a large extent on investment costs, precisely tailored processes, adequate competencies, and the acceptance of employees. Industry 4.0 is therefore not a purely technical challenge, but has an equal impact on technologies, organization and human beings. Against the backdrop of this triad, the consortial project addresses a holistic risk management in the introduction of Industrie 4.0.

Project management: Prof. Dr.-Ing. Roman Dumitrescu, Heinz Nixdorf Institute of Paderborn University

Data-driven retrofit and product generation planning in mechanical and plant engineering (DizRuPt)

Duration: 01/2019 - 06/2022
Total project volume: 2.3 million euros
Project volume of the University: 661,000 euros
Funded by: Federal Ministry of Education and Research

The research project DizRuPt enables companies to exploit the use phase data of their products and other relevant data from the product life cycle in strategic product planning. By systematically analyzing these data, new features and functions can be derived. This enables the usage-oriented planning of retrofits as well as future product generations. The intended solution focuses on:

  • Methods for data acquisition and analysis
  • Methods for deriving new product functions for retrofit and generation planning
  • Organizational implementation through suitable processes, structures, and competencies
  • IT tools to support the organizational implementation

Project management: Prof. Dr.-Ing. Roman Dumitrescu, Heinz Nixdorf Institute of Paderborn University

Rapidly explainable artificial intelligence for industrial plants (RAKI)

Duration: 01.09.2019 - 19.06.2022
Total project volume: 1.9 million euros
Project volume of the University: 414,543 euros
Funded by: Federal Ministry for Economic Affairs and Energy

RAKI develops novel methods for scalable, comprehensible machine learning with "humans in the loop".

The project focuses on scalable AI-driven optimization of the configuration and operation of industrial plants as well as necessary production logistics.

Distributed implementations enable the processing of large amounts of data for the automatic generation of explanations. The development and application partners AI4BD and Siemens plan to use key parts of the RAKI framework after integration into their CBR and Mindsphere platforms. The results of RAKI form the basis for novel data products such as AI-driven interactive configuration software for industrial plants, enabling scalable development of smart services in industrial production.

Project management: Prof. Dr. Axel-Cyrille Ngonga Ngomo, Paderborn University

Project partners: AI4BD Deutschland GmbH, Siemens AG, Leipzig Universität

Junior research group „Data Driven Methods in Control Engineering”

Duration: 01.07.2020 bis 30.06.2024
Total project volume (University): 1,591,507 euros
Funded by: Federal Ministry of Education and Research

In the course of digitalization, artificial intelligence and machine learning are currently receiving a great deal of attention from science and industry. In control engineering, data-driven methods are already used, but mainly as an alternative to physical modeling of dynamic behavior or to subject-specific methods of control design or in a pragmatic simple combination.

Therefore, the goal of the junior research group "DART - Data Driven Methods in Control Engineering" is to develop novel hybrid methods for control engineering problems by combining the well-established physically motivated methods with modern data driven methods to achieve the highest possible performance in control design. These hybrid approaches go far beyond simple, pragmatic combinations because they are based on structurally well-justified compositions of methods tailored to each other that synergistically combine their advantages. The typical design steps such as modeling and parameter identification of the physical system, observer design, controller design and commissioning of the controller are addressed. Thus, we are able to extend all aspects of classical control engineering as a complete system by hybrid approaches with data-based methods.

Project management: Dr.-Ing. Julia Timmermann, Heinz Nixdorf Institute of Paderborn University

Explainable Diagnostic AI for Industrial Data (DAIKIRI)

Duration: 2020 - 2022
Total project volume: 1.4 million euros
Project volume of the University: 500,00 euros
Funded by: Federal Ministry of Education and Research

The DAIKIRI research project aims to develop the first automatic procedures for the semantification of industrial data and data-driven diagnosis of industrial plants. These methods will be used to develop diagnostic smart services for industrial data and to evaluate them with data from real use cases. DAIKIRI will develop AI procedures that are self-explanatory and automatically verbalize AI results, making them more transparent. Users will thus be able to understand how the results were obtained and decisions based on these results can be made with confidence.

Project management: Prof. Dr. Axel-Cyrille Ngonga Ngomo, Paderborn University

Project partners: USU Software AG, elevait GmbH & Co. KG, pmOne AG

Simultaneous Development and Testing of Cyber Physical Systems (CPS) using the example of autonomous electric vehicles (SET CPS)

Duration: 2019 - 2022
Total project volume: 1 million euros
Project volume of the University: 482,000 euros
Funded by: European Regional Development Fund

Paderborn/Aachen, May 28, 2019. How can autonomous vehicles with electric drives be developed as examples of complex cyber-physical systems faster, more cost-effectively, and with lower resource consumption? And how can the safety of these vehicles on the road be increased? A team of researchers and developers from dSPACE, e.GO Mobile AG, and the Institute of Industrial Mathematics at the University of Paderborn started a research project a few weeks ago to answer this complex question. The project is funded by the German state of North Rhine-Westphalia (NRW) and the EU as part of the IKT.NRW lead market competition. The project, scheduled to run for 36 months, aims to simultaneously develop and test cyber-physical systems (CPS) using the example of an electric autonomous vehicle. It is abbreviated SET CPS according to its German title.

In vehicle development, trends such as automated driving and the development of alternative drives, such as battery-powered vehicles, are causing a sharp increase in the demands placed on the underlying systems. When these types of vehicles are developed, the aim is to optimize a large number of target parameters such as fuel consumption, range, and driving comfort, and to guarantee the safety of the system. Researchers and developers in the SET CPS project are now looking for new approaches to make the development processes for manufacturers and suppliers reliable and economical, and enable them to meet development times.

The project therefore aims to develop intelligent, simulation-based processes that improve and systematize the development and test process of complex vehicles and increase the degree of automation. For this purpose, design and testing are more closely interlinked to achieve a high level of quality even in the early development phases. The researchers also use the latest mathematical methods from multi-objective optimization, which is one of the core competencies of the Institute of Industrial Mathematics. This enables them to simultaneously achieve competing goals, such as energy efficiency, comfort, and costs, while ensuring the safety of the system. The plan is to integrate the new processes into the dSPACE tool chain and evaluate them using an example from e.GO vehicle development.

“As consortium leader of the project, our goal is to take the next step toward a one-stop development environment for autonomous vehicles," explained Dr. Rainer Rasche, Group Manager Test Automation at dSPACE. “The resulting tool chain enables the developer to adjust the parameters of an ECU to different, typical traffic situations and simultaneously test them in the simulated environments. This will enable our customers to accelerate their development."

Dr. Michael Riesener, Vice President Corporate Research at e.GO Mobile AG, said: “The simultaneous development and testing of new systems for our electric vehicles made possible by SET CPS also enables us to achieve fast development times and to design the vehicles with an even stronger focus on requirements. For this reason, we look forward to advancing the research project in cooperation with our partners."

About e.GO Mobile AG

e.GO Mobile AG was founded in 2015 by Prof Dr Günther Schuh as a manufacturer of electric vehicles. The more than 400 employees use the campus's unique network of research facilities and approximately 360 technology companies on the RWTH Aachen Campus. Highly agile teams work on a variety of cost-effective and customer-focused electric vehicles for short-haul traffic. e.GO Mobile AG is currently commissioning its new plant in Rothe Erde, Aachen, for series production.

Project management: Dr. Sebastian Peitz, Paderborn University

Project partners: Prof. Dr. Michael Dellnitz, Paderborn University; dSPACE; e.GO Mobile AG

FutureLab Power Electronics – Integrated Lab for future Wide-Bandgap-based Power Electronic Applications enabling highest Miniaturization & Efficiency

Duration: 01.01.2019 – 30.06.2022
Total project volume (University):  944.132 euros
Funded by: Federal Ministry of Education and Research - BMBF (Initiative "Research Laboratories Microelectronics Germany" (ForLab))

Motivation

Wide bandgap (WBG) power semiconductors like Gallium-Nitride (GaN) or Silicon-Carbide (SiC) show massive innovation potential for various power electronic applications. These new semiconductor technologies characterized by significantly reduced switching losses enable both, power electronic systems with highest efficiency and outstanding power densities, allowing a considerable miniaturization of power electronic applications. As a consequence, also the system costs of a large number of applications, like e.g. On-Board Chargers and DC-DC-Converters of automotive EVs, IT-Power Supplies of data centers and mobile telecom networks (5G and beyond), Renewable Energy generation/transmission/distribution, Medical Appliances (CT, MRT, ultrasonics) can be decreased, although the WBG-semiconductor components today show higher costs than their conventional, silicon-based counterparts. Next to the component costs, the comparably rare reliability data of the new WBG-power semiconductors are considered as a last hurdle to be passed before ramping up an industrial large-scale utilization of this groundbreaking technology. This ramp up, moreover would lead to reduced WBG-component costs further accelerating the technology transition. 

Project objectives

Accordingly, two main objectives will be addressed at LEA within the next years: Identification, development and optimization of further advantageous applications benefiting from WBG-based power designs as well as improving reliability of those WBG-based power electronic applications. Focusing on these topics, special support shall be offered also to small- and medium-sized enterprises within dedicated follow-up projects.

The new lab infrastructure currently being acquired and put into operation in our 2-story lab building (IW) will enable LEA to adequately meet these objectives throughout the next decade.

Project strategy

The FUTURE LAB equipment enables the following technological steps as strategical subjects:

  • Reducing the size of chokes, transformers and filter components by drastically increasing the switching frequency of power electronic
  • Characterization of new WBG-semiconductors regarding switching behaviour and losses
  • Validation of core and winding losses in magnetic components
  • Professional manufacturing of custom-made mechanical components like coil formers or ferrite cores for specialized magnetics
  • Extensive qualification of developed demonstrators and prototypes, incl.:
  • EMI testing and optimization
  • Testing and improving operation under harsh climate conditions
  • Reliability analysis of components and prototype systems under climatic stress conditions

Technological Challenges

Below challenges are considered and will be addressed by appropriate lab equipment:

  • Fast switching of WBG-semiconductors leads to a high demand in EMI-filtering and -shielding
  • High dv/dt-rates stress the isolation of transformers, drive circuits and auxiliary power supplies
  • Novel packages of power semiconductors require special concepts for cooling and new thermal solutions

Envisaged Investments

The main laboratory equipment investments are:

  • Professional CAD-tool for multilayer PCB-design (SW)
  • Precise assembly unit for smallest SMD-components
  • 3D-printer for specifically designed mechanical components
  • CNC-milling machine for application-optimized ferrite cores within magnetic components
  • Coil winding machine (CNC) for magnetic components
  • Testbed for high-dv/dt switching loss measurements
  • Testbed for calorimetric efficiency measurements
  • Testbed for EMI-measurements
  • Climatic exposure test chamber with vibrating table for environmental testing

Project management: Prof. Dr.-Ing. Joachim Böcker, Paderborn University

Project partners: Fraunhofer IISB Erlangen, Fraunhofer ENAS Paderborn

Adaptive control of nonlinear differential-algebraic systems in multibody dynamics

Duration: 2017 - 2022
Total project volume: 595.592 euros
Project volume of the University: 307.146 euros
Funded by: German Research Foundation (DFG)

Our objective in the proposed project is to develop adaptive controller design techniques for tracking control of systems of nonlinear differential-algebraic equations with applications to underactuated mechanical multibody systems. While closed-loop tracking control of fully actuated multibody systems is well-established, systematic methods for underactuated systems are lacking. The latter type refers to multibody systems having more degrees of freedom than actuators, which results in diverse systems theoretic properties. Typical examples of practical relevance are systems with passive joints, cranes, cable robots or lightweight systems with flexible bodies. Especially for more complex systems with kinematic loops or flexible bodies, differential-algebraic equations are appropriate for modelling. In the proposed project, we first aim to conduct a structural analysis of multibody systems. Thereby it is intended to characterize important systems theoretic quantities and properties such as input-to-state stability, index, relative degree and internal dynamics on the basis of physically motivated considerations. Problems in controller design for multibody systems may arise when the index or relative degree of the differential-algebraic model exceed one or the system has unstable internal dynamics. To compensate a higher relative degree, the funnel observer, which has been developed by the applicants Berger and Reis, shall be applied. Unstable internal dynamics are aimed to be circumvented by an application of feedforward control strategies based on model inversion. Such a model inversion shall be based on so-called servo-constraints, which again lead to differential-algebraic equations. The performance and implementability of the developed methods is to be constantly verified by means of selected experiments.

Project management: Jun.-Professor Dr. Thomas Berger at Paderborn University

Cryptography-Based Security Solutions: Enabling Trust in New and Next Generation Computing Environments (Sub-project E1 at Collaborative Research Center 1119, CROSSING)

Duration: 07/2018 - 06/2022
Total project volume (Sub-project E1): 516.200 euros
Project volume of the University: 302.500 euros
Funded by: German Research Foundation (DFG)

Within the DFG's Collaborative Research Center (Sonderforschungsbereich) 1119, CROSSING, we are heading the project on the Secure Integration of Cryptographic Software.

Together with Mira Mezini's Software Technology Group, we are researching means to aid developers in integrating cryptographic libraries securely into their software systems in sub-project E1.

Project management: Prof. Eric Bodden (sub-project E1.), Heinz Nixdorf Institute at Paderborn University

Speaker of the SFB1119: Prof. Marc Fischlin, TU Darmstadt

Intelligent user support for vulnerability analysis (IntelliScan)

Duration: 15.09.2017 - 15.04.2022
Total project volume (University): 451.329 euros
Funded by: Ministry for Culture and Science of the State of North Rhine-Westphalia

Funding is provided for researchers in IT security at various universities and colleges in North Rhine-Westphalia on the topic of "Human Centered Systems Security". IntelliScan is one of the five funded research tandems. It focuses on vulnerability analysis.

Current automated tools for vulnerability analysis are often difficult to use and, moreover, cannot be extended or are difficult to extend, making it difficult to apply them to new contexts of use. The aim of the "IntelliScan" project is therefore to explore new types of user support for such tools and to evaluate and further develop them by means of user studies. As a result, tools for vulnerability analysis should be easy to use and also extendable for common developers. In this way, programme analysis is to be transformed from a technology for specialists into a standard technique.

Project management: Prof. Dr. Eric Bodden, Heinz Nixdorf Institute at Paderborn University and Prof. Dr. Matthew Smith, University Bonn

Future-proofing the Soot Framework for Programme Analysis and Transformation (FutureSoot)

Duration: 01.03.2018 – 28.02.2022
Total project volume (University): 395.400 euros
Funded by: German Research Foundation (DFG)

Soot is the world’s most popular framework for analysing and transforming Java and Android programs. Over its more than twenty-year lifespan, countless scientific tools have emerged that build directly on Soot. The Soot framework hereby provides these tools with a common implementation platform, which greatly increases the comparability of the individual approaches, and greatly accelerates the implementation of the individual tools. The project FutureSoot aims to put the Soot framework on the right track in order to guarantee its maintenance for a long time beyond the project funding. For this reason, the project includes the development of a sustainability concept, the establishment and further expansion of a reliable build-and-test infrastructure, as well as further work on greater modularisation of the core components. The work is designed to make Soot easier to maintain in the future, to maintain it according to a proven and well-documented plan, and to bring together the main stakeholders of the Soot community through workshops and coordinate further maintenance and development among them.

Project management: Prof. Dr. Eric Bodden, Heinz Nixdorf Institute of Paderborn University and Professor Dr. Rüdiger Kabst, Paderborn University

Distributed Acoustic Signal Processing over Wireless Sensor Networks

Duration: 2020 - 2023
Total project volume (University): 283,500 euros
Funded by: German Research Foundation (DFG)

The project "Distributed Acoustic Signal Processing over Wireless Sensor Networks" conducts research on signal processing anc communication over acoustic sensor networks. Such an acoustic sensor network consists of nodes with microphones (possibly also loudspeakers) and communicates using wireless transmission techniques. The sensor network is typically connected to the Internet via one or more gateways. Such a network carries out acoustic applications (such as noise reduction and speaker separation). An obvious approach would be to transport all acoustic data collected by the sensor nodes to the gateway for processing. However, this is not necessarily the best possible approach (because of the high communication load, possibly large latency). It also gives away the potential benefit of processing the data already on the sensor nodes, thus reducing data volumes and latency. We are therefore developing approaches (inspired by trends such as microservices and network function virtualization) where acoustic signal processing can be broken into individual processing blocks and distributed over the individual nodes. In this follow-on project, we intend to continue our work on both distributed acoustic signal processing and the development of a framework for automated distribution of such blocks over the network. Specifically, we are looking at hardware aspects (especially with full-duplex audio capabilities) and the possibilities of using such hardware for synchronization in time-based MAC protocols or for estimating acoustic round trip times. This allows us to estimate and appropriately calibrate the spatial geometry of the setup in both static and dynamic environments. Building on information about the acoustic utility of individual nodes, we will address the problem of source selection (signals from which microphones should be included?) both from an  acoustic and a network perspective. A key question will be how to handle such network aspects in dynamic scenarios. Dynamicity arise in particular from movement of acoustic sources and recording devices; such movements can be uncontrolled or controlled (e.g. in the case of robotic sensor nodes). Our work will result in a  testbed that integrates the essential functions and will be an experimental environment for practical testing of the developed algorithms.

Coordination project

In daily life, we are surrounded by a multitude of noises and other acoustic events. Nevertheless we are able to effortlessly converse in such an environment, retrieve a desired voice while disregarding others, or draw conclusions about the composition of the environment and activities therein, given the observed sound scene. A technical system with similar capabilities would find numerous applications in fields as diverse as ambient assisted living, personal communications, and surveillance. With the continuously decreasing cost of acoustic sensors and the pervasiveness of wireless networks and mobile devices, the technological infrastructure of wireless acoustic sensor networks is available, and the bottleneck for unleashing new applications is clearly on the algorithmic side.

This Research Unit aims at rendering acoustic signal processing and classification over acoustic sensor networks more 'intelligent', more adaptive to the variability of acoustic environments and sensor configurations, less dependent on supervision, and at the same time more trustworthy for the users. This will pave the way for a new array of applications which combine advanced acoustic signal processing with semantic analysis of audio. The project objectives will be achieved by adopting a three-layer approach treating communication and synchronization aspects on the lower, signal extraction and enhancement on the middle, and acoustic scene classification and interpretation on the upper layer. We aim at applying a consistent methodology for optimization across these layers and a unification of advanced statistical signal processing with Bayesian learning and other machine learning techniques. Thus, the project is dedicated to pioneering work towards a generic and versatile framework for the integration of acoustic sensor networks in several classes of state-of-the-art and emerging applications.

Sound Recognition with Limited Supervision over Sensor Networks

A fundamental problem for many machine learning methods is a discrepancy between the training data and the test data in a later application, which can lead to a significant drop in the classification rate. In acoustic event detection and scene classification in acoustic sensor networks, this problem is exacerbated because there is a very large number of possible sounds and because such networks can be deployed in widely varying geometric configurations and environments. For this reason, existing databases for acoustic event and scene classification will virtually never be a perfect fit for a new application in an acoustic sensor network.

The main goal of this project is therefore to develop methods that allow existing databases to be usable for distinct audio classification tasks in an acoustic sensor network despite this discrepancy. We assume that weakly annotated data, i.e., annotated only with the event class but not with timestamps, are available from another domain, and that unannotated data is available from the target domain. Procedures are now being developed to compute a strong annotation from a weak annotation in which start and end times of acoustic events are additionally annotated to compute domain invariant features, as well as procedures to perform domain adaptation to overcome differences between training data and test scenario in this way. We are also investigating adaptation at test time to adapt to changing acoustic environments and sensor configurations. In particular, deep generative neural models will be used. Appropriate network structures and objective functions will be developed to separate the different factors influencing the observed waveform, in particular the variation caused by the acoustic event from the variation of the signal caused by the environment. Furthermore, we will develop methods to detect unusual acoustic events, because these can be of particular importance for a distinct application.

Project management: Dr.-Ing. Jörg Schmalenströer, Paderborn University and Prof. Dr. Holger Karl, Hasso-Plattner-Institut 

Algorithms for Programmable Matter in a Physiological Medium (PROGMATTER)

Duration: 2018 - 2022
Total project volume: 259.139 euros
Project volume of the University: 200.206 euros
Funded by: German Research Foundation (DFG)

The goal of this project is the development of new models and algorithms for swarms of autonomous nano robots, i.e. groups of tiny active units with very limited computational capabilities that act as a swarm in a physiological medium.

For the development of appropriate models for these robot swarms, we will consider questions that have not or have only been partly addressed in the past: How can nano robots generate, store, and use energy? Which models would be reasonable for non-local communication between a group of nano robots? Which assumptions would be sufficiently realistic for the failure and repair of nano robots by themselves or other nano robots?

It has turned out that the leader election problem represents a key problem in order to come up with solutions for many other important problems for swarms of nano robots. Therefore, we will check whether the models developed within this project will allow us to solve the leader election problem efficiently. Based on appropriate solutions, we will then investigate solutions for other important problems: (1) Explore an area by a swarm of nano robots and potentially mark important points. (2) Coat an object by a swarm of nano robots as quickly as possible. (3) Search for a near-shortest path for a swarm of nano robots to a particular target point.

Besides looking for upper bounds, we will also investigate lower bounds for the efficiency of solutions within our models.

Finally, we plan to extend our algorithmic results by aspects like robustness to failures of nano robots and solutions in 3-dimensional space.

Project management: Prof. Dr. Christian Scheideler, Paderborn University, Germany and Prof. Shlomi Dolev, Ben Gurion University of the Negev, Israel

Intelligent assistance system to support the digital transformation of business models – Smart GM

Duration: 01/2020 - 12/2022
Total project volume: 1,4 million euros
Project volume of the University: 836,000 euros
Funded by: European Union and State of North Rhine-Westphalia

The ability to develop innovative business models for one’s own products and services is of central importance for every company. At the same time, however, many small and medium-sized enterprises (SMEs) in particular find it difficult to fill the abstract term ʺbusiness model innovation“ with life, i. H. developing business model innovations in a targeted and systematic manner. This increases the risk that innovative products and services will not be successfully marketed – which in turn harms the competitiveness of companies and thus endangers jobs and social prosperity. This is exactly where this project comes in. In the Smart GM project at the SICP – Software Innovation Campus Paderborn, the SI-Lab of the University of Paderborn, the professorships Kundisch, Hüllermeier and Wünderlich and the companies myconsult, UNITY, WP Kemper and Fellowmind are working together on an assistance system that suggests its users suitable innovative business model ideas. The basis for this is, on the one hand, an extensive knowledge base on business models and, on the other hand, artificial intelligence. The AI-algorithms should generate new ideas from the large number of possible combinations. These are then evaluated on a public crowd platform and by customers and experts. With an increasing number of ratings, the quality of new business model proposals of the assistance systems is also increased.

For the first time, the Smart GM project combines competencies and methods from the areas of business model innovations, technology acceptance, machine learning, (crowd-based) evaluation of idea quality and computer-aided idea generation for the development of business model innovations and paves the way for a new generation of business model innovation methods – from passive support to active assistance.

Project management: Prof. Dr. Dennis Kundisch, Paderborn University

TheaterLytics

Duration: 01.06.2019 - 31.05.2022
Total project volume: 652,000 euros
Project volume of the University: 414,000 euros
Funded by: State of North Rhine-Westphalia for the promotion of digital model regions, in accordance with the circular of the Ministry for Economic Affairs, Innovation, Digitalisation and Energy of the State of North Rhine-Westphalia.

Good cultural offerings are not the only guarantee for the success of a cultural institution. Many decisions have to be made every day that affect visitor satisfaction, but also the economic situation of the business. Most often this happens on personal experience. With the TheaterLytics project, a decision support system for data-based revenue management and the offer design is to be developed for the first time.

In particular, forecasting the utilization of cultural events is a major challenge for event planners. Too many influencing factors are playing a role here: holidays, weather, weekdays and also the genre are just a few factors that cannot be considered simultaneously. The project provides event planners with an IT tool for targeted planning.

A decision support system (DSS) for data-based revenue management and the offer design is conceived and implemented as a prototype into software. Both visitors and non-visitors are questioned to create a database. Building on this, methods and models for capacity utilization forecasting are then developed in order to ensure better planning of resources and capacities. Methods and models for event scheduling, hall management and pricing are also being developed.

Project management: Prof. Dr. Dennis Kundisch, Paderborn University

Agile teamwork through predictive competence management (PredicTeams)

Duration: 10/2020 - 09/2023
Total project volume: 2.4 million euros
Project volume of the University: 456,000 euros
Funded by: Ministry for Economic Affairs, Innovation, Digitalisation and Energy of the State of North Rhine-Westphalia

The PredicTeams project aims to develop a practice-oriented framework for predictive competence management for agile teams. In order to enable companies to manage the transition to agile teamwork in digital work environments, the project focuses on the following goals:

  •  Identification and operationalization of competences for agile teamwork in the context of digital working environments and the provision of a database with instruments for measuring relevant competences
  • Method for simplifying the process of competence assessment using semantic language analysis
  • Methodology for the analysis of competence profiles
  • Show-case applications for assessing competences with the help of semantic text analysis and for the analysis of competence profiles on the basis of fuzzy-set qualitative comparative analysis
  • Models and methods as well as a guideline for the design of a predictive competence management and future-oriented team staffing

The goals are achieved by taking up state-of-the-art measuring instruments and methods in the field of human resources and organizational research as well as empirical methodology, developing them further, adapting them for use in companies and testing them based on test data. For this purpose, employees’ competence data are evaluated and newly collected in order to significantly reduce assessment dimensions. Exemplary data analyses will be implemented and evaluated. In addition to the identification of the most important competences, the project aims to simplify the process of competence assessment through spoken comments and automated text analysis. The simplified process allows for continuous assessments and moves away from annual written statements. The planned measures will provide the basis for a time-efficient, state-of-the-art assessment and analysis of competences in companies. Through the use of the developed methods and models, companies are enabled to move away from administrative competence management and towards predictive competence management. The project is both practically and scientifically innovative, since predictive HR analytics is much discussed, but is still in its infancy.

The project is conducted in the context of the technology cluster it's OWL. It is supported by the Ministry for Economics, Innovation, Digitalization and Energy of NRW. The project is funded with 2.4 mio. Euro.

Project management: Prof. Dr. Kirsten Thommes, Paderborn University

Flexible monitoring and control systems for the energy and mobility transition in the distribution grid using artificial intelligence (FLEMING)

Duration: 01.09.2019 - 31.08.2022
Total project volume: 5,861,703 euros
Project volume of the University: 571,336 euros
Funded by: Federal Ministry for Economic Affairs and Energy, Funding code: 03EI6012F

The FLEMING project aims to revolutionize continuous function monitoring, specifically the current use of sensors in distribution grids, by combining artificial intelligence (AI) methods with advancements in sensor technology, and thus contribute significantly to the success of Germany's energy and mobility transition.

The focus of German climate and energy policy is on a massive and area-wide integration of plants for the generation of renewable energies as well as on an integration of charging stations for electromobility into the existing power grid. The resulting numerous load fluctuations - e.g., caused by decentralized solar plants - as well as the temporally and spatially concentrated energy demand caused by charging infrastructure (eMobility) lead to a very large load on electrical equipment and components, up to overload. At the same time, grid operators are under increasing efficiency and cost pressure.

Critical Relevance of the Current Network Condition

Grid operators require, on the one hand, a better understanding of the current status of the existing grid and its components in order to meet the goals of the energy and mobility transition while keeping the same quality of supply (monitoring). This will enable for the early detection and prediction of possible damage and equipment failures, as well as the prevention of such failures through improved control. Smart load management, on the other hand, necessitates the use of sensors that are sufficiently accurate, reliable, and easy to retrofit. Only then will more flexible grid utilization, which takes advantage of temporary overload potential, be possible, as well as the nationwide expansion of energy distribution infrastructure that will be required in the future, especially in the light of the rapidly increasing electrification of the automotive sector.

The scenario calls for the end-to-end use of sensors, information and communication systems to collect the necessary data from the individual network resources and components. Sensor solutions for condition monitoring that are currently available are only used in niche or peripheral applications. A universal application fails at the moment due to very complicated engineering as well as the sensor systems' limited life span and performance, limiting them to simple monitoring tasks, mostly of individual equipment. Furthermore, existing sensor technology is typically only available for one manufacturer's plants, preventing transferability and obstructing generic, system-wide data processing. The goal of the project is to improve present sensor deployment in distribution grids by combining artificial intelligence (AI) methodologies with sensor technology expansion. All key features of sensor deployment in electrical equipment are included in the sub-goals derived from this.

Project management: Prof. Dr. Daniel Beverungen (Paderborn University) and Prof. Dr. Eyke Hüllermeier (Paderborn University)

Project partners: ABB AG Forschungszentrum Deutschland, Forschungsinstitut für Rationalisierung e.V. (FIR) from Aachen, Karlsruher Institut für Technologie (KIT), SÜC Energie and H2O GmbH from Coburg and Heimann Sensor GmbH

Multimodal culture platform “OWL Live”

Duration: 01.01.2020 - 31.12.2022
Total project volume: 1.1 million euros
Project volume of the University: 242,824 euros
Funded by: European Union and the State of North Rhine-Westphalia

The interactive, multimodal culture platform “OWL Live” intends to bundle the cultural offerings of the Ostwestfalen-Lippe (OWL) region, make them more visible and usable in the future, as well as establish as many interfaces as possible to existing systems. OWL.LIVE is for actors and public service providers, cultural intermediaries, and citizens. The goal is threefold. First, we strive to enable citizens to find suitable cultural events through individualized filter mechanisms. Second, we will better connect cultural stakeholders across sectors, strengthen the visibility of voluntary workers and associations, overcome regional borders, and – especially for OWL as a rural area – guarantee sufficient mobility to enable cultural participation. Third, actors and public service providers benefit from “OWL Live” as it provides services for organizing cultural events and other projects. In summary, OWL.LIVED is an intelligent, target group-specific, user-oriented platform that contributes to establishing OWL as a cultural brand, making the cultural audience perceive the rich cultural region OWL more completely.

Project management: Prof. Dr. Daniel Beverungen, Paderborn University

Project partners: OstwestfalenLippe GmbH, Bielefeld; aXon Gesellschaft für Informationssysteme mbH, Paderborn

EcoDrive

Duration: 01.01.2018 - 31.12.2022
Total project volume: not published due to contractual conditions
Project volume of the University: not published due to contractual conditions​​​​​​​
Funded by: Cottbus Chamber of Industry and Commerce, several automotive companies, telematics companies and logistics companies

The transportation sector accounts for approximately 25% of the Greenhouse Gas emissions. Minimizing fuel consumption is one of the main goals of climate actions. For logistic companies, minimizing fuel consumption has the beneficial side effect of lowering variable costs. Even though social and firm goals are aligned, many firms fail to bring down fuel consumption. One of the main reasons is the driving behavior of truck drivers.

In this project, we combine insights from behavioral science & technology and improve feedback mechanisms for truck drivers. We use feedback mechanisms and gamification. Our first results led to a sustainable reduction of about 10% of the fuel.

We also analyze how environmental conditions, especially varies aspects of traffic density affect drivers’ inclination towards automation and eventually good driving behavior.

Project management: Prof. Dr. Kirsten Thommes, Paderborn University

Project partners: Brandenburg University of Technology and various industry partners

Remaining Useful Lifetime for New and Used Technical Systems under Non-Stationary Conditions (REASON)

Duration: April 2021 till March 2024
Total project volume: 603.220 euros
Project volume of the University: 307.686 euros
Funded by: German Research Foundation (DFG)

The project aims to develop methods for predicting the remaining useful life (RUL) of systems operating under non-stationary conditions, such as varying loads and speeds. Therefore, classical empirical models from the engineering field will be combined with methods from the field of artificial intelligence. This hybrid approach will be used to categorize operating conditions, identify failure modes, and predict the RUL of technical systems. In addition, these methods will be employed to predict the RUL of systems already in use that have been retrofitted with suitable sensors but where no sensor data from their past operation is available.

Project management: Prof. Dr.-Ing. habil. Walter Sextro, Paderborn University and Prof. Dr. Eyke Hüllermeier, Ludwig-Maximilians-Universität München

Project partners: Prof. Dr. Eyke Hüllermeier, Ludwig-Maximilians-Universität München

Funnel MPC with application to the control of magnetic levitation systems

Duration: 2022 - 2025
Total project volume: 455.554 euros
Project volume of the University: 226.260 euros
Funded by: German Research Foundation (DFG)

The objective of the proposed research project is the development, numerical implementation and analysis of the new control concept Funnel MPC (FMPC). This concept ties adaptive tracking control, learning and optimization based methods together in an innovative way. Funnel control and model predictive control (MPC) are both current research areas in control engineering and mathematical systems theory, which successfully balance theory and application. FMPC utilizes known advantages of both control strategies (e.g., compliance with output and control constraints, inherent robustness, excellent control performance) to achieve the long-term goal of a universal controller design for nonlinear systems. FMPC consists of three components:

1.) In a model-based part of the controller, elements from funnel control are integrated into MPC, e.g., by incorporating the high-gain factor from the funnel controller in the construction of the stage costs. This ensures compliance with the output constraints and ultimately allows to rigorously prove recursive feasibility via an optimality argument – without (stabilizing) terminal constraints and independent of the length of the prediction horizon.

2.) MPC does not guarantee robustness in general. Hence, it is a main objective to transfer the robustness inherent to funnel control to FMPC. To this end, the control loop is extended by a model-free component via coupling with a funnel controller with respect to the prediction error of the model-based part. For this combination, robustness with respect to model uncertainties is to be proved rigorously.

3.) Through a second extension of the control loop by a learning component a continuous model adaptation as well as a concomitant improvement of controller performance is achieved. For this purpose, unknown model parameters are approximated and the system state is estimated. Meanwhile, the robustified FMPC guarantees the strict satisfaction of the output constraints. Additionally, as numerical tests have shown, it induces a sufficient stimulation of the system, which ensures a high information content in the input-output data that is necessary for the learning process. This is to be characterized in a mathematically rigorous and laid out in a verifiable way by the concept of „persistency of excitation“ within the project. As a proof of concept the control of magnetic levitation trains will be considered, where a regular feedback between theory and numerical practice is intended. In levitation control a prescribed distance between vehicle suspension and guideway must be ensured. Furthermore, a robustness with respect to uncertainties (e.g., the total mass of the vehicle depending on the occupancy of the passenger area) and disturbances (e.g., wind conditions) is crucial. At the same time, a high controller performance, including travelling comfort, is desirable. Exactly those properties are unified in the innovative concept FMPC.

Project management: Jun. Prof. Thomas Berger at Paderborn Unviersity

Training, validation and benchmark tools for the development of data-driven operating and control processes for intelligent, local energy systems (DARE)

Duration: 10/2021 - 09/2023
Total project volume (University): 994.000 euros
Funded by: Federal Ministry of Education and Research - BMBF

With the funding approval of the Federal Ministry of Education and Research (BMBF), the project “Training, validation and benchmark tools for the development of data-driven operating and control processes for intelligent, local energy systems” (DARE) started at the beginning of October 2021. Over the next two years, scientists from the SCIP – Software Innovation Campus Paderborn will develop an open-source simulation and benchmark framework together with scientists from the Competence Center for Sustainable Energy Technology (KET) and the associated economic partners WestfalenWIND GmbH and Westfalen Weser Netz GmbH. The framework is intended to address problems that can arise when operating decentralized energy networks. The overarching goal of the project is to promote the transformation of the current energy supply system to a sustainable structure characterized by renewable energies.

Microgrids as a solution component for the energy transition

The transformation towards a sustainable, efficient and cost-effective energy supply structure is one of the central challenges of the 21st century. In order to realize the energy transition, cellular and decentralized energy systems, so-called microgrids, can represent an important solution component. Microgrids are local energy networks that operate both grid-connected and autonomously in stand-alone operation and can supply industrial companies and households with energy. They consist of energy sources (e.g. wind turbines), energy stores (e.g. batteries) and energy consumers from different sectors (electricity, heat, mobility).

“Microgrids have the advantage that, thanks to their local integration, renewable energy can be made available close to where it is consumed and can therefore be used directly by the consumer over a short distance. As a result, national energy networks can be relieved and the need for network expansion decreases. In addition, the proportion if regenerative energies increases, since the lossy transport over long distances and unnecessary shutdowns of regenerative power plants due to grid bottlenecks are avoided”, explains Dr. Gunnar Schomaker, R&D Manager “Smart Systems” at SCIP.

Central component for the production of the basic energy supply in emerging and developing countries

“The fact that microgrids can also operate autonomously in island mode is a typical case for remote, off-grid areas. In addition to the contribution to the energy transition in Europe, the microgrid represents a central building block for the production of the basic energy supply in emerging and developing countries (especially Sub-Saharan Africa), since the development of a central energy infrastructure in sparsely populated, rural areas is not feasible in the long term”, explains Dr.-Ing. Oliver Wallscheid, scientific director of the research project.

Challenges in operating microgrids

Microgrids can bring great potential for the energy transition and the establishment of the basic energy supply in emerging and developing countries, but this is also accompanied by challenges that still have to be overcome. The main challenge, and thus also the central research question of the project, is to ensure a consistent and efficient energy supply through operating and control processes. “Compared to the classic, central large networks, there are challenges with decentralized networks that affect stability, among other things. Because a secure energy supply is much more difficult to maintain in decentralized grids than in central grids, which are supported by conventional large power plants, due to the volatility of regenerative power plants and typically only low storage and reserve capacities”, explains Dr. Wallscheid.

“The traditional top-down strategies of large central networks cannot therefore be transferred to the operation and control of such stochastic, heterogeneous and volatile energy networks”, says Jun. Prof. Dr. Sebastian Peitz. “Instead, data-driven and self-learning processes are emerging as a possible solution, e.g. from the field of reinforcement learning. However, the problem here is that these learning and innovative control methods cannot be used directly in the field due to safety and availability aspects, but must first be improved and evaluated on the basis of synthetic data in a closed simulation cycle”, adds Jun. Prof. Peitz.

Although there are already solutions, they are also very heterogeneous and are often based on greatly simplified model environments, so that no statements can be made about a future transfer to practice. In addition, there is no establishment comparison standard that can be used to objectively and quantifiably evaluate data-driven controllers.

Open source simulation and benchmark framework

“The goal within our DARE project is therefore to build an open-source simulation and benchmark framework that maps the previously explained problem framework when operating decentralized energy networks. The research into data-driven controllers for energy technology should be accelerated and made comparable through easily accessible and standardized training, validation and benchmark tools”, says Dr. Wallscheid.

Through the integration of economic partners from energy technology practice, the project also attaches great importance to the depiction of realistic evaluation scenarios. The open source framework to be created will therefore also make an important contribution to the transfer of data-driven controllers from simulation to field use.

Project management: Dr.-Ing. Oliver Wallscheid, Paderborn University and Jun.-Prof. Dr. Sebastian Peitz, Paderborn University

Project partners: Kompetenzzentrum für nachhaltige Energietechnik (KET); Software Innovation Lab (SI-Lab); WestfalenWIND GmbH; Westfalen Weser Netz GmbH; Prof. Dr. Eyke Hüllermeier, LMU München

Process Mining for the Analysis and Prescription of industrial core processes (BPM-I4.0)

Duration: 01.04.2021 - 31.03.2023
Total project volume: 1.743.292 euros
Project volume of the University: 387.120 euros
Funded by: Ministry for Economic Affairs, Innovation, Digitalisation and Energy of the State of NRW

Processes form the organizational core of companies and help to structure them. Process mining can be applied to analyze these processes in a data-driven way. This approach is already established in business sectors such as online retailing. But industrial processes like product creation or individual order processing of machines require a high degree of creativity and expertise. Process mining in this field has not been sufficiently researched neither in practice or in science. This can be explained by a lack of availability of sufficient amounts of data on such processes. In addition, the data sets of industrial processes are more unstructured and flexible than e.g., an order process at a mail order company for standard goods, which makes the application of process mining much more difficult and addresses different challenges for the analysis.

The project "BPM-I4.0" aims at a complete development, implementation, and evaluation of process mining techniques for the mentioned industrial processes. This includes the analysis of past and running process instances, as well as the prediction of future process steps and the providing of targeted recommendations for action through prescriptive methods. For this purpose, innovative methods, concepts, algorithms and digital tools are developed and prototypically applied and evaluated in the product development process of Weidmüller GmbH & Co KG in Detmold and the order processing process of GEA Westfalia Separator Group GmbH in Oelde. CONTACT Software is also involved as a development partner to contribute its expertise in process mining and product lifecycle management. Paderborn University is represented by the SI-Lab of the Software Innovation Campus Paderborn with the departments of Prof. Daniel Beverungen and Prof. Oliver Müller. Also researchers of the Fraunhofer IEM are involved in the project.

The results of this project will support companies to improve their core processes by analyzing process data and proactively managing the execution of their processes to stay in competition in the medium and long term. Furthermore, the project provides important, scientific results in the still young research field of prescriptive process mining. In addition, the active application in the business environment also offers the opportunity to develop relevant contributions.

Project management: Prof. Dr. Daniel Beverungen, Paderborn University

Project partners: GEA Westfalia Separator Group GmbH, Oelde; CONTACT Software GmbH, Paderborn; Weidmüller GmbH & Co KG, Detmold; Fraunhofer Institute for Mechatronic Systems Design (IEM), Paderborn

Netzwerk „SustAInable Life-cycle of Intelligent Socio-Technical Systems“ (SAIL)

Laufzeit: 01.08.2022 - 31.07.2026
Fördervolumen gesamt: 16,40 Mio. Euro
Fördervolumen der Universität: 5,5 Mio. Euro
Gefördert durch: Ministerium für Kultur und Wissenschaft des Landes Nordrhein-Westfalen

SAIL adressiert die nächste Stufe der KI-Entwicklung, indem der gesamte Lebenszyklus von KI-Systemen und deren technologische und gesellschaftliche Auswirkungen in den Blick genommen werden. SAIL ist dementsprechend interdisziplinär angelegt und bindet Wissenschaftler*innen aus der Kern-KI, aus den Ingenieurwissenschaften sowie aus den Sozial- und Geisteswissenschaften ein. Das Forschungsprogramm ist inhaltlich in drei Grundlagensäulen („research pillars“) und zwei Anwendungsgebiete („application domains“) unterteilt. In der Grundlagenforschung werden zum einen das Zusammenspiel von KI und menschlichen Partnern bei der Bewertung und Abstimmung von Fehlern und Zielen betrachtet. Außerdem werden ausgereifte KI-Systeme analysiert, um deren möglicherweise unerwünschte langfristige Auswirkungen auf funktionaler, kognitiver und gesellschaftlicher Ebene zu modellieren und abzumindern. Zuletzt wird der gesamte KI-Lebenszyklus im Hinblick auf Effizienz betrachtet, damit der praktische Einsatz von KI-Systemen mit möglichst wenig Energie-, Zeit- und Speicherbedarf und geringer kognitiver Anstrengung beim menschlichen Partner ermöglicht wird.

Die Anwendungsgebiete von SAIL sind intelligente industrielle Arbeitsumgebungen und adaptive Assistenzsysteme für die Gesundheitsfürsorge.

Projektleitung: Prof. Dr. Axel-Cyrille Ngonga-Ngomo, Institut für Informatik der Universität Paderborn

Projektpartner: Wissenschaftler`*innen der Universität Paderborn (Prof. Dr. Katharina Rohlfing, Jun. Prof. Dr. Ilona Horwath, Jun. Prof. Dr. Suzana Alpsancar, Prof. Dr. Eric Bodden, Prof. Dr. Roman Dumitrescu, Prof. Dr. Reinhold Häb-Umbach, Jun. Prof. Dr. Sebastian Peitz, Prof. Dr. Marco Platzner, Prof. Dr. Christian Plessl, Prof. Dr. Walter Sextro, Prof. Dr. Ansgar Trächtler

Contact
Phone:
+49 5251 60-6277
Fax:
+49 5251 60-6297
Office:
F0.328
Web:

Google Autocomplete Challenge with Prof. Dr. Axel-Cyrille Ngonga Ngomo on "artificial intelligence". Unfortunately, the video is only available in German.

Interdisciplinary research facilities

Heinz Nixdorf Institute

Interdisciplinary Research Institute of the Paderborn University

 

DaSCo

Paderborn Institute for Data Science and Scientific Computing 

 

PC²

Paderborn Center for Parallel Computing

 

KET

Competence Centre for Sustainable Energy Technology

 

SI-Lab

Software Innovation Lab

 

IFIM

Institute for Industrial Mathematics

 

JAII

Joint Artificial Intelligence Institute

 

Cooperation partners

it's OWL

Technology Network it's OWL - Intelligent Technical Systems OstWestfalenLippe

 

SICP

Software Innovation Campus Paderborn

 

Fraunhofer IEM

Fraunhofer Institute for Mechatronic Systems Design

 

L-LAB

Institute for Automotive Lighting Technology and Mechatronics

 

The University for the Information Society