Produktion und Systeme
Refine
Year of publication
- 2020 (4) (remove)
Document Type
Language
- English (4)
Is part of the Bibliography
- no (4)
Keywords
- Activity recognition (1)
- Applied Research (1)
- Collaboration (1)
- Intelligent robots (1)
- Service robots (1)
- Synchronization (1)
- composite material (1)
- magnetoelectric effect (1)
- magnetostriction (1)
- piezoelectric effect (1)
Institute
- Fakultät Elektro- und Informationstechnik (4) (remove)
Begutachtungsstatus
- peer-reviewed (3)
The ability to synchronize expectations among human-robot teams and understand discrepancies between expectations and reality is essential for human-robot collaboration scenarios. To ensure this, human activities and intentions must be interpreted quickly and reliably by the robot using various modalities. In this paper we propose a multimodal recognition system designed to detect physical interactions as well as nonverbal gestures. Existing approaches feature high post-transfer recognition rates which, however, can only be achieved based on well-prepared and large datasets. Unfortunately, the acquisition and preparation of domain-specific samples especially in industrial context is time consuming and expensive. To reduce this effort we introduce a weakly-supervised classification approach. Therefore, we learn a latent representation of the human activities with a variational autoencoder network. Additional modalities and unlabeled samples are incorporated by a scalable product-of-expert sampling approach. The applicability in industrial context is evaluated by two domain-specific collaborative robot datasets. Our results demonstrate, that we can keep the number of labeled samples constant while increasing the network performance by providing additional unprocessed information.
A voltage transformer employing the magnetoelectric effect in a composite ceramic heterostructure with layers of a magnetostrictive nickel–cobalt ferrite and a piezoelectric lead zirconate–titanate is described. In contrast to electromagnetic and piezoelectric transformers, a unique feature of the presented transformer is the possibility of tuning the voltage transformation ratio K using a dc magnetic field. The dependences of the transformer characteristics on the frequency and the amplitude of the input voltage, the strength of the control magnetic field and the load resistance are investigated. The transformer operates in the voltage range between 0 and 112 V, and the voltage transformation ratio K is tuned between 0 and 14.1 when the control field H changes between 0 and 6.4 kA/m. The power at the transformer output reached 63 mW, and the power conversion efficiency was 34%. The methods for calculation of the frequency response, and the field and load characteristics of the transformer are proposed. The ways to improve performance characteristics of magnetoelectric transformers and their possible application areas are discussed.
Thermomechanical shape memory materials have certain disadvantages when it comes to 3D volumetric reproduction intended for rapid prototyping or robotic prehension. The need to constantly supply energy to counteract elastic retraction forces in order to maintain the required geometry, together with the inability to achieve conformal stability at elevated temperatures, limits the application of thermal shape memory polymers. Form removal also presents problems as most viscoelastic materials do not ensure demolding stability. This work demonstrates how magnetoactive boron−organo−silicon oxide polymers under the influence of an applied magnetic field can be used to achieve energy free sustainable volumetric shape memory effects over extended periods. The rheopectic properties of boron−organo−silicon oxide materials sustain form removal without mold distortion.
The Applied Research Conference which is held every year at another University of Applied Sciences in Bavaria is the main event for all students in the Master of Applied Research program. They come together to present their work in oral presentations and full papers, which are published in the proceedings, as well as poster presentations of the students after their 1st semester. For sure it is interesting for Professors and interested people to see the results of upcoming scientists and researchers. Due to the Corona pandemic, this year it is not possible to organize the conference as usual in presence. We have to refrain from face-to-face discussions, having together a cup of coffee. We as Professors at the OTH Regensburg wanted to give our students the chance to finish this semester successfully – despite all limitations due to the pandemic situation. Therefore we decided to organize the RARC 2020 – Regensburg Applied Research Conference 2020 – for the Master of Applied Research students of OTH Regensburg as an online conference. However, for a Technical University the situation is rather a challenge than a problem. Using a variety of online tools for teaching during this semester, we have enough experience to find a setup for RARC2020. We received 28 submissions for full papers, which were peer reviewed and 26 of them were accepted – you will find them in this proceedings, and they will be presented orally on July 31st, 2020. Additionally, there are 22 posters which will be presented on the same day. During the Plenary Opening Session, after a welcome by our President Prof. Dr. Wolfgang Baier we will have 3 Keynote speakers:
Prof. Dr. phil. habil. Karsten Weber: Erkenntnistheorie für Ingenieure
Prof. em. Georg Scharfenberg: 11 Years Master of Applied Research Alumnis
Veronika Fetzer: Entrepreneurship