Home  | Legals | Data Protection | Sitemap | KIT

Dipl.-Ing. Martin Krause

Research Associate
department: Machines, Equipment and Process Automation
office hours: to be agreed
room: 012, Geb. 50.36
phone: +49 1523 9502592
Martin KrauseBdu0∂kit edu

76131 Karlsruhe
Kaiserstraße 12

Dipl.-Ing. Martin Krause

Area of Research:

  • Expert in machine learning (Supervised/ Semi-Supervised/ Unsupervised Learning)
  • Using reinforcement learning to enhance the productivity of production systems
  • Modal analysis; vibration analysis
  • Dynamics of robot systems (multibody simulation)
  • Configuration of robot systems


General Tasks:




Test benches:


Dissertation: Increasing productivity of sorting processes with reinforcement learning


Curriculum Vitae:

since 05/2014 Research Associate at the Institute of Production Science (wbk) at Karlsruhe Institute of Technology (KIT)
10/2007 - 04/2014 Study of Mechanical Engineering at Karlsruhe Institute of Technology (KIT)



[ 1 ] Sell-Le Blanc, F.; Fleischer, J.; Li, L. & Krause, M. (2014), "Characterization of Process-Machine-Interaction for Coil Winding Processes using Multi Body Dynamics". CIRP Conference on Virtual Machining Process Technology, eds. Elsevier, pp. 1-8.
Coil winding processes for electric motors offer an opportunity in production to boost the performance of single tooth winding and their layer technique for the use in electric drives. Single tooth winding, also known as concentrated winding, consists of separate coil bobbins which are assembled to a stator after the winding process. The most suitable manufacturing process for single teeth, with respect to speed and product quality, is the already established linear winding process. Given a higher geometry aspect ratio of the bobbin, due to the use in electric drives, wire tension control efforts rise caused by an increased impact of wire deformation effects. Lower productivity and higher commissioning efforts are the drawbacks. In order to meet these new challenges a detailed characterization of the process-machine-interaction is necessary. Given this background, a new innovative approach that considers the wire material influences and describes the interaction between the elastoplastic wire deformation, machine parameters as well as process-parameters will be presented. The combined approach to link the mechanical model using rigid bodies with a material-based model using elastic bodies has the potential to exceed the accuracy given by conventional approaches. This paper comprises an overview of different simulation aspects for the chained multi-body joint wire model as a core feature. The model application in form of a case study will demonstrate the ability for further process optimization.

[ 2 ] Krause, M.; Bauer, J. & Fleischer, J. (2015), "Zuführung kleiner Bauteile über Schwingförderer", wt Werkstattstechnik online, pp. 627-632.
Die effiziente Teilezuführung ist eine herausfordernde Handhabungsoperation in der automatisierten Mikromontage. Für die Zuführung von kleinen Bauteilen unterschiedlichster Art ist vor allem das Prinzip der Gleitförderung eine interessante Alternative. Um mit diesem Prinzip eine Produktivitätssteigerung zu erreichen, ist ein Ansatz notwendig, mit dem sich die Bauteilbewegung vorhersagen lässt.

[ 3 ] Fleischer, J. & Krause, M. (2016), "Separation strategies for a feeding system with modular coupled piezoelectric vibration conveyers". Proceedings of the 16th International Conference of the European Society for Precision Engineering and Nanotechnology, eds. EUSPEN, pp. 195-196.
Microassembly is characterised by a large variety of small parts. An approach to supply those variety of different parts with precise and gentle movements is the principle of sliding conveyance. Detected by a vision system, the parts can be separated automatically, sorted and precisely positioned underneath the handling device with independently operating piezoelectric vibration conveyors, which form a kind of panel-carpet. The challenge here is represented by parts showing, different position-dependent feeding velocities and therefore a part specific behaviour. However, these varying feeding velocities can be predicted by a double staged model, which comprises the process of sliding conveyance as well as the behaviour of the phase flexible piezoelectric conveyor system. Based on the predicted velocities of the fed parts, the article outlines an optimisation procedure utilising evolutionary algorithms to determine the optimal separation strategies. The separation process is achieved by individually switched on feeding units of dedicated panels of the conveyor system, an optimal choice of feeding directions as well as by selected process parameters for the sliding conveyance process describing the target function. A further constraint is required by straighten out parts at the junction of the feeding units to ensure reliability of the separation process.

[ 4 ] Hofmann, J.; Sell-Le Blanc, F.; Krause, M.; Wirth, F. & Fleischer, J. (2016), "Simulation of the Assembly Process of the Insert Technique for Distributed Windings". Proceedings of 6th International Electric Drives Production Conference (E|DPC), eds. IEEE, pp. 144-148.
As the efficiency of electric power trains in hybrid electric vehicles should be increased and at the same time the manufacturing costs reduced, different motor designs and production concepts need to be considered. Because of the nearly sinusoidal magnetic field inside, the stator design with distributed windings, which is typically produced with the insert technique, is technically preferred. The insert technique offers a high productivity because the complete winding assembling process can be done in one step. This results in the fact, that nearly 80% of all electric motors worldwide have distributed windings. In order to enhance the possibilities for distributed windings with the insert technique, the current fill factor needs to be improved. Due to the fact that the actual wire placement cannot be measured und thus not optimized, a simulative approach with a multi-body simulation is used to understand the process interactions between the wires, the stator groove and the tool. This approach will be presented in this paper.