Refine
Department, Institute
- Fachbereich Wirtschaftswissenschaften (904)
- Fachbereich Informatik (792)
- Fachbereich Angewandte Naturwissenschaften (476)
- Institut für funktionale Gen-Analytik (IFGA) (434)
- Fachbereich Elektrotechnik, Maschinenbau, Technikjournalismus (353)
- Institute of Visual Computing (IVC) (251)
- Institut für Technik, Ressourcenschonung und Energieeffizienz (TREE) (230)
- Fachbereich Sozialpolitik und Soziale Sicherung (207)
- Internationales Zentrum für Nachhaltige Entwicklung (IZNE) (65)
- Institut für Sicherheitsforschung (ISF) (59)
Document Type
- Article (1482)
- Conference Object (1276)
- Part of a Book (664)
- Book (390)
- Report (117)
- Contribution to a Periodical (54)
- Doctoral Thesis (53)
- Lecture (44)
- Conference Proceedings (41)
- Master's Thesis (38)
Year of publication
Language
Has Fulltext
- no (4302) (remove)
Keywords
- Lehrbuch (69)
- Deutschland (24)
- Controlling (23)
- Unternehmen (22)
- Nachhaltigkeit (20)
- Management (19)
- Betriebswirtschaftslehre (16)
- Sozialversicherung (14)
- Corporate Social Responsibility (13)
- Prozessmanagement (13)
Optimization plays an essential role in industrial design, but is not limited to minimization of a simple function, such as cost or strength. These tools are also used in conceptual phases, to better understand what is possible. To support this exploration we focus on Quality Diversity (QD) algorithms, which produce sets of varied, high performing solutions. These techniques often require the evaluation of millions of solutions -- making them impractical in design cases. In this thesis we propose methods to radically improve the data-efficiency of QD with machine learning, enabling its application to design. In our first contribution, we develop a method of modeling the performance of evolved neural networks used for control and design. The structures of these networks grow and change, making them difficult to model -- but with a new method we are able to estimate their performance based on their heredity, improving data-efficiency by several times. In our second contribution we combine model-based optimization with MAP-Elites, a QD algorithm. A model of performance is created from known designs, and MAP-Elites creates a new set of designs using this approximation. A subset of these designs are the evaluated to improve the model, and the process repeats. We show that this approach improves the efficiency of MAP-Elites by orders of magnitude. Our third contribution integrates generative models into MAP-Elites to learn domain specific encodings. A variational autoencoder is trained on the solutions produced by MAP-Elites, capturing the common “recipe” for high performance. This learned encoding can then be reused by other algorithms for rapid optimization, including MAP-Elites. Throughout this thesis, though the focus of our vision is design, we examine applications in other fields, such as robotics. These advances are not exclusive to design, but serve as foundational work on the integration of QD and machine learning.
Current knowledge about cell-biomaterial interactions is often based on 2D cell culture systems like protein-coated glass slides. However, such smooth surfaces cannot mimic the nanofibrous environment of the native extracellular matrix (ECM). It is therefore a major challenge to transfer the results from 2D surfaces to 3D protein scaffolds with biomimetic nanofiber architecture. To understand the influence of different protein topographies on the cell response we introduce a new process to fabricate binary collagen scaffolds of variable thickness with spatially controlled regions of nanofibrous and smooth topography. We used pH-induced self-assembly to prepare collagen nanofibers with diameters between 130 and 150 nm on glass surfaces, which were partly covered with a polymer mask. After cross-linking with glutaraldehyde, smooth collagen films were prepared on the remaining glass regions. Atomic force microscopy revealed a much lower surface roughness of smooth collagen compared to nanofibers.
When a robotic agent experiences a failure while acting in the world, it should be possible to discover why that failure has occurred, namely to diagnose the failure. In this paper, we argue that the diagnosability of robot actions, at least in a classical sense, is a feature that cannot be taken for granted since it strongly depends on the underlying action representation. We specifically define criteria that determine the diagnosability of robot actions. The diagnosability question is then analysed in the context of a handle manipulation action, such that we discuss two different representations of the action – a composite policy with a learned success model for the action parameters, and a neural network-based monolithic policy – both of which exist on different sides of the diagnosability spectrum. Through this comparison, we conclude that composite actions are more suited to explicit diagnosis, but representations with less prior knowledge are more flexible. This suggests that model learning may provide balance between flexibility and diagnosability; however, data-driven diagnosis methods also need to be enhanced in order to deal with the complexity of modern robots.
Toshiyuki Fukao
(2020)
Quantitative estimates for the global impact of COVID-19 on the diagnosis and management of patients with inborn errors of metabolism (IEM) are lacking. We collected relevant data from 16 specialized medical centers treating IEM patients in Europe, Asia and Africa. The median decline of reported IEM related services in March 1st-May 31st 2020 compared to the same period in 2019 were as high as 60-80% with a profound impact on patient management and care for this vulnerable patient group. More representative data along with outcome data and guidelines for managing IEM disorders under such extraordinary circumstances are needed.
