Deep Concrete Flow: Deep learning based characterisation of fresh concrete properties from open-channel flow using spatio-temporal flow fields

authored by: Max Coenen, Christian Vogel, Tobias Schack, Michael Haist
Abstract: The global production of concrete, being one of the most commonly utilised materials in the world, is associated with significant CO2-emissions and a substantial depletion of mineral resources. In order to improve sustainability, concretes thus are increasingly produced using recipes containing a large variety of different raw materials, including e.g. CO2 reduced cements or recycled materials and industrial wastes. However, these actions result in heightened susceptibility of the concrete to variations in raw material characteristics, consequently diminishing the concrete’s resilience and decreasing the concrete’s robustness. Against this background, the quality control of fresh concrete before casting becomes of significant importance. However, current quality control is mainly conducted based on analogous, empirical, and often subjective test approaches. In this paper, a novel method is introduced for automatically and comprehensively evaluating the quality of fresh concrete at construction sites. In particular, the paper presents an end-to-end trainable framework for the image-based characterisation of fresh concrete properties. More specifically, a camera-based setup is proposed, in which image sequences of the discharge process of a mixing truck are acquired, on the basis of which the fresh concrete properties such as the consistency and rheology are determined. In this context, this paper introduces the concept of Spatio-Temporal Flow Fields, which provide a compact representation of the concrete’s flow behaviour and serve as input to a multi-task convolutional neural network (CNN) predicting the target parameters describing the fresh concrete’s properties. A thorough examination of the performance of the suggested method is carried out using highly challenging real-world data, showcasing extremely compelling outcomes with average prediction errors for the concrete properties of only about 6%.
Organisation(s): Institute of Building Materials Science
Type: Article
Journal: Construction and Building Materials
Volume: 411
ISSN: 0950-0618
Publication date: 12.01.2024
Publication status: Published
Peer reviewed: Yes
ASJC Scopus subject areas: Materials Science(all), Building and Construction, Civil and Structural Engineering
Electronic version(s): https://doi.org/10.1016/j.conbuildmat.2023.134809 (Access: Closed)

BibTeX

@article{874dd08e2f664b6fa614d2c59f16b257,
title = "Deep Concrete Flow: Deep learning based characterisation of fresh concrete properties from open-channel flow using spatio-temporal flow fields",
abstract = "The global production of concrete, being one of the most commonly utilised materials in the world, is associated with significant CO2-emissions and a substantial depletion of mineral resources. In order to improve sustainability, concretes thus are increasingly produced using recipes containing a large variety of different raw materials, including e.g. CO2 reduced cements or recycled materials and industrial wastes. However, these actions result in heightened susceptibility of the concrete to variations in raw material characteristics, consequently diminishing the concrete{\textquoteright}s resilience and decreasing the concrete{\textquoteright}s robustness. Against this background, the quality control of fresh concrete before casting becomes of significant importance. However, current quality control is mainly conducted based on analogous, empirical, and often subjective test approaches. In this paper, a novel method is introduced for automatically and comprehensively evaluating the quality of fresh concrete at construction sites. In particular, the paper presents an end-to-end trainable framework for the image-based characterisation of fresh concrete properties. More specifically, a camera-based setup is proposed, in which image sequences of the discharge process of a mixing truck are acquired, on the basis of which the fresh concrete properties such as the consistency and rheology are determined. In this context, this paper introduces the concept of Spatio-Temporal Flow Fields, which provide a compact representation of the concrete{\textquoteright}s flow behaviour and serve as input to a multi-task convolutional neural network (CNN) predicting the target parameters describing the fresh concrete{\textquoteright}s properties. A thorough examination of the performance of the suggested method is carried out using highly challenging real-world data, showcasing extremely compelling outcomes with average prediction errors for the concrete properties of only about 6%.",
keywords = "Fresh concrete, Deep learning, Computer vision, Open-channel flow, Spatio-temporal flow fields, Concrete quality control, Rheology",
author = "Max Coenen and Christian Vogel and Tobias Schack and Michael Haist",
note = "The authors acknowledge the funding of the project ReCyCONtrol (https://www.recycontrol.uni-hannover.de/en/) provided by the German Federal Ministry of Education and Research (BMBF) under the grant No. 033R260 A and the funding of the project Open Channel Flow provided by the German Research Foundation (DFG) under the grant No. 452024049. We furthermore acknowledge the project Safe Concrete Pumping – Pumpability and Pumping Stability of Concrete funded by the German Federal Ministry for Economic Affairs and Climate Action (BMWK) under the grant No. 20947 BG, in the course of which parts of the experimental data used in this paper were acquired. The data used in this paper can by requested by contacting the corresponding author and may only be used for research purposes.",
year = "2024",
month = jan,
day = "12",
doi = "10.1016/j.conbuildmat.2023.134809",
language = "English",
volume = "411",
journal = "Construction and Building Materials",
issn = "0950-0618",
publisher = "Elsevier Ltd.",
}