Details
| Titel in Übersetzung | AI-supported quality assurance for the flow production of UHFB bar elements |
|---|---|
| Originalsprache | Deutsch |
| Seiten (von - bis) | 34-41 |
| Seitenumfang | 8 |
| Fachzeitschrift | Beton- und Stahlbetonbau |
| Jahrgang | 116 |
| Ausgabenummer | S2 |
| Publikationsstatus | Veröffentlicht - 6 Sept. 2021 |
Abstract
AI-supported quality assurance for the flow production of UHFB bar elements. Fast precision construction requires modular designs with lightweight, easy-to-handle components that can be manufactured at high repetition rates. To enable such elements to be produced continuously, quickly and economically, new manufacturing and design concepts are needed, in which the design must be highly efficient (form follows force) and which also consistently complies to the requirements for rapid series production. AI-supported quality assurance controls in the course of flow production, which obtains its data from a variety of different sensors and sensor types, lend themselves to such new manufacturing processes. Such AI-assisted manufacturing is implemented by an Artificial Neural Network (KNN), which, compared to conventional approaches, offers the possibility to map nonlinear relationships of heterogeneous data. In this paper, the KNN-based form of quality assurance (QA) on flow-produced beam elements is described as an example for the assessment of the concrete surface and geometry, which can represent a first step into process control. First investigations have shown that even small defects can be detected and localized very efficiently.
Schlagwörter
- artificial neural networks, flow production, machine learning, quality assurance, ultra high performance concrete
ASJC Scopus Sachgebiete
- Ingenieurwesen (insg.)
- Bauwesen
Zitieren
- Standard
- Harvard
- Apa
- Vancouver
- BibTex
- RIS
in: Beton- und Stahlbetonbau, Jahrgang 116, Nr. S2, 06.09.2021, S. 34-41.
Publikation: Beitrag in Fachzeitschrift › Artikel › Forschung › Peer-Review
}
TY - JOUR
T1 - KI-gestützte Qualitätssicherung für die Fließfertigung von UHFB-Stabelementen
AU - Tritschel, Franz Ferdinand
AU - Markowski, Jan
AU - Penner, Nikolai
AU - Rolfes, Raimund
AU - Lohaus, Ludger
AU - Haist, Michael
PY - 2021/9/6
Y1 - 2021/9/6
N2 - AI-supported quality assurance for the flow production of UHFB bar elements. Fast precision construction requires modular designs with lightweight, easy-to-handle components that can be manufactured at high repetition rates. To enable such elements to be produced continuously, quickly and economically, new manufacturing and design concepts are needed, in which the design must be highly efficient (form follows force) and which also consistently complies to the requirements for rapid series production. AI-supported quality assurance controls in the course of flow production, which obtains its data from a variety of different sensors and sensor types, lend themselves to such new manufacturing processes. Such AI-assisted manufacturing is implemented by an Artificial Neural Network (KNN), which, compared to conventional approaches, offers the possibility to map nonlinear relationships of heterogeneous data. In this paper, the KNN-based form of quality assurance (QA) on flow-produced beam elements is described as an example for the assessment of the concrete surface and geometry, which can represent a first step into process control. First investigations have shown that even small defects can be detected and localized very efficiently.
AB - AI-supported quality assurance for the flow production of UHFB bar elements. Fast precision construction requires modular designs with lightweight, easy-to-handle components that can be manufactured at high repetition rates. To enable such elements to be produced continuously, quickly and economically, new manufacturing and design concepts are needed, in which the design must be highly efficient (form follows force) and which also consistently complies to the requirements for rapid series production. AI-supported quality assurance controls in the course of flow production, which obtains its data from a variety of different sensors and sensor types, lend themselves to such new manufacturing processes. Such AI-assisted manufacturing is implemented by an Artificial Neural Network (KNN), which, compared to conventional approaches, offers the possibility to map nonlinear relationships of heterogeneous data. In this paper, the KNN-based form of quality assurance (QA) on flow-produced beam elements is described as an example for the assessment of the concrete surface and geometry, which can represent a first step into process control. First investigations have shown that even small defects can be detected and localized very efficiently.
KW - artificial neural networks
KW - flow production
KW - machine learning
KW - quality assurance
KW - ultra high performance concrete
UR - http://www.scopus.com/inward/record.url?scp=85114337784&partnerID=8YFLogxK
U2 - 10.1002/best.202100052
DO - 10.1002/best.202100052
M3 - Artikel
VL - 116
SP - 34
EP - 41
JO - Beton- und Stahlbetonbau
JF - Beton- und Stahlbetonbau
SN - 0005-9900
IS - S2
ER -