• German

Main Navigation

Collaborative Research Center SFB 876 - Providing Information by Resource-Constrained Data Analysis

The collaborative research center SFB876 brings together data mining and embedded systems. On the one hand, embedded systems can be further improved using machine learning. On the other hand, data mining algorithms can be realized in hardware, e.g. FPGAs, or run on GPGPUs. The restrictions of ubiquitous systems in computing power, memory, and energy demand new algorithms for known learning tasks. These resource bounded learning algorithms may also be applied on extremely large data bases on servers.

Resource-aware Machine Learning - Virtual International Summer School 2020

Machine learning has become one of the driving fields in data analysis. But how to deal with data analysis and limited resources: Computational power, data distribution, energy or memory? The summer school on Resource-aware Machine Learning provides lectures on the latest research in machine learning, typically with the twist on resource consumption and how these can be reduced. This year’s summer school will be held online and free of charge between 31st of August and 4th of September. The events will be a mixture of pre-recorded and live sessions, including a dedicated space for presenting PhD/PostDoc research and a hackathon featuring real world ML tasks.


A selection of course topics: Deep Learning, Graph Neural Networks, Large Models on Small Devices, Power Consumption of ML, Deep generative modeling, Memory challenges in DNN...

More info and registration: https://www-ai.cs.tu-dortmund.de/summer-school-2020/

During registration you may express your interest in joining the hackathon and/or presenting at the students’ corner.


Hackathon - Positioning prediction and robot control

As a practical example to leverage your ML skills we host a challenge on indoor location prediction based on floor-integrated sensor data. Real world data is gathered in a warehouse scenario, where free roaming robots perform transportation of goods. Your first task will be to use sensor data (vibration, magnetic fields…) with position ground truth to build a position prediction. The best teams will get the chance for live control of the robots on the final day of the summer school. The winner will be invited for research cooperations to Dortmund.

More details: https://www-ai.cs.tu-dortmund.de/summer-school-2020/hackathon

Students’ Corner - Share and discuss your work

The summer school will be accompanied by an exchange platform for participants, the Students' Corner, which will allow them to network and share their research. During the registration you may express your interest in participation at the student’s corner and we will keep you updated.

More details: https://www-ai.cs.tu-dortmund.de/summer-school-2020/students-corner

The summer school is organised by the competence center for Machine Learning Rhine-Ruhr, ML2R, the collaborative research center SFB 876 and the artificial intelligence group at TU Dortmund University.

mehr ...

New Article in Engineering Applications of Artificial Intelligence - Download Now!

We are happy to announce that the newest article of project B3 "Real-time prediction of process forces in milling operations using synchronized data fusion of simulation and sensor data" is now available at ScienceDirect (follow: this link) for free access until August, 9th.

The paper focuses on Machine Learning based predictions in milling processes. In mechanical engineering, milling is one of the most important machining operations with a wide variety of application use cases, e.g., the machining of structural components for the aerospace industry, dental prostheses or forming tools in the context of the tool and die manufacturing. Different challenges arise for different process strategies, milling tools and machine tools, such as tool vibrations of long and slender finishing tools causing chatter marks on the workpiece surface and tool wear for long-running processes.

Nowadays, in the context of Industry 4.0, Machine Learning methods have allowed production processes, including machining, to be better understood and intelligently transformed. Data gathered and evaluated during these processes is the fundamental basis of such transformation. The industrial processes can thus not only be better understood, but also optimized.

In this context, to prevent undesirable effects during milling processes, the paper proposes a novel approach for combining simulation data with sensor data to generate online predictions of process forces, which are influenced by tool wear, using an ensemble-based machine learning method. In addition, a methodology was developed in order to synchronize pre-calculated simulation data and streaming sensor measurements in real time. Sensor data was acquired using milling machines by the Virtual Machining group of the Chair for Software Engineering in the laboratories of the Institute of Machining Technology (ISF), TU Dortmund. The geometric physically-based simulation system has also been developed in the same chair.

mehr ...

Show news archive
Newsletter RSS Twitter