Searcharxiv⌕ Search

arXiv subjects

Daniel Schall

Publications and source records attributed to Daniel Schall.

At least 19 recordsLinked to original sources
arXiv2023

Comparison of Clustering Algorithms for Statistical Features of Vibration Data Sets

Vibration-based condition monitoring systems are receiving increasing attention due to their ability to accurately identify different conditions by capturing dynamic features over a broad frequency range. However, there is little research on clustering approaches in vibration data and the resulting solutions are often optimized for a single data set. In this work, we present an extensive comparison of the clustering algorithms K-means clustering, OPTICS, and Gaussian mixture model clustering (GMM) applied to statistical features extracted from the time and frequency domains of vibration data sets. Furthermore, we investigate the influence of feature combinations, feature selection using principal component analysis (PCA), and the specified number of clusters on the performance of the clustering algorithms. We conducted this comparison in terms of a grid search using three different benchmark data sets. Our work showed that averaging (Mean, Median) and variance-based features (Standard Deviation, Interquartile Range) performed significantly better than shape-based features (Skewness, Kurtosis). In addition, K-means outperformed GMM slightly for these data sets, whereas OPTICS performed significantly worse. We were also able to show that feature combinations as well as PCA feature selection did not result in any significant performance improvements. With an increase in the specified number of clusters, clustering algorithms performed better, although there were some specific algorithmic restrictions.

cs.LG↗
arXiv2023

Federated Learning for Predictive Maintenance and Quality Inspection in Industrial Applications

Data-driven machine learning is playing a crucial role in the advancements of Industry 4.0, specifically in enhancing predictive maintenance and quality inspection. Federated learning (FL) enables multiple participants to develop a machine learning model without compromising the privacy and confidentiality of their data. In this paper, we evaluate the performance of different FL aggregation methods and compare them to central and local training approaches. Our study is based on four datasets with varying data distributions. The results indicate that the performance of FL is highly dependent on the data and its distribution among clients. In some scenarios, FL can be an effective alternative to traditional central or local training methods. Additionally, we introduce a new federated learning dataset from a real-world quality inspection setting.

cs.LG↗
arXiv2022

Autoencoder based Anomaly Detection and Explained Fault Localization in Industrial Cooling Systems

Anomaly detection in large industrial cooling systems is very challenging due to the high data dimensionality, inconsistent sensor recordings, and lack of labels. The state of the art for automated anomaly detection in these systems typically relies on expert knowledge and thresholds. However, data is viewed isolated and complex, multivariate relationships are neglected. In this work, we present an autoencoder based end-to-end workflow for anomaly detection suitable for multivariate time series data in large industrial cooling systems, including explained fault localization and root cause analysis based on expert knowledge. We identify system failures using a threshold on the total reconstruction error (autoencoder reconstruction error including all sensor signals). For fault localization, we compute the individual reconstruction error (autoencoder reconstruction error for each sensor signal) allowing us to identify the signals that contribute most to the total reconstruction error. Expert knowledge is provided via look-up table enabling root-cause analysis and assignment to the affected subsystem. We demonstrated our findings in a cooling system unit including 34 sensors over a 8-months time period using 4-fold cross validation approaches and automatically created labels based on thresholds provided by domain experts. Using 4-fold cross validation, we reached a F1-score of 0.56, whereas the autoencoder results showed a higher consistency score (CS of 0.92) compared to the automatically created labels (CS of 0.62) -- indicating that the anomaly is recognized in a very stable manner. The main anomaly was found by the autoencoder and automatically created labels and was also recorded in the log files. Further, the explained fault localization highlighted the most affected component for the main anomaly in a very consistent manner.

cs.LG↗
arXiv2022

Smart Active Sampling to enhance Quality Assurance Efficiency

We propose a new sampling strategy, called smart active sapling, for quality inspections outside the production line. Based on the principles of active learning a machine learning model decides which samples are sent to quality inspection. On the one hand, this minimizes the production of scrap parts due to earlier detection of quality violations. On the other hand, quality inspection costs are reduced for smooth operation.

cs.LG↗
arXiv2022

Two-dimensional Platinum Diselenide Waveguide-Integrated Infrared Photodetectors

Low cost, easily integrable photodetectors (PDs) for silicon (Si) photonics are still a bottleneck for photonic integrated circuits (PICs), especially for wavelengths above 1.8 $μ$m. Multilayered platinum diselenide (PtSe$_2$) is a semi-metallic two-dimensional (2D) material that can be synthesized below 450$°$C. We integrate PtSe$_2$ based PDs directly by conformal growth on Si waveguides. The PDs operate at 1550 nm wavelength with a maximum responsivity of 11 mA/W and response times below 8.4 $μ$s. Fourier transform infrared spectroscopy (FTIR) in the wavelength range from 1.25 $μ$m to 28 $μ$m indicates the suitability of PtSe$_2$ for PDs far into the infrared wavelength range. Our PtSe$_2$ PDs integrated by direct growth outperform PtSe$_2$ PDs manufactured by standard 2D layer transfer. The combination of IR responsivity, chemical stability, selective and conformal growth at low temperatures, and the potential for high carrier mobility, make PtSe$_2$ an attractive 2D material for optoelectronics and PICs.

physics.app-ph↗
arXiv2021

Evaluation of Hyperparameter-Optimization Approaches in an Industrial Federated Learning System

Federated Learning (FL) decouples model training from the need for direct access to the data and allows organizations to collaborate with industry partners to reach a satisfying level of performance without sharing vulnerable business information. The performance of a machine learning algorithm is highly sensitive to the choice of its hyperparameters. In an FL setting, hyperparameter optimization poses new challenges. In this work, we investigated the impact of different hyperparameter optimization approaches in an FL system. In an effort to reduce communication costs, a critical bottleneck in FL, we investigated a local hyperparameter optimization approach that -- in contrast to a global hyperparameter optimization approach -- allows every client to have its own hyperparameter configuration. We implemented these approaches based on grid search and Bayesian optimization and evaluated the algorithms on the MNIST data set using an i.i.d. partition and on an Internet of Things (IoT) sensor based industrial data set using a non-i.i.d. partition.

cs.LG↗
arXiv2021

Minimal-Configuration Anomaly Detection for IIoT Sensors

The increasing deployment of low-cost IoT sensor platforms in industry boosts the demand for anomaly detection solutions that fulfill two key requirements: minimal configuration effort and easy transferability across equipment. Recent advances in deep learning, especially long-short-term memory (LSTM) and autoencoders, offer promising methods for detecting anomalies in sensor data recordings. We compared autoencoders with various architectures such as deep neural networks (DNN), LSTMs and convolutional neural networks (CNN) using a simple benchmark dataset, which we generated by operating a peristaltic pump under various operating conditions and inducing anomalies manually. Our preliminary results indicate that a single model can detect anomalies under various operating conditions on a four-dimensional data set without any specific feature engineering for each operating condition. We consider this work as being the first step towards a generic anomaly detection method, which is applicable for a wide range of industrial equipment.

cs.LG↗
arXiv2021

Towards Robust and Transferable IIoT Sensor based Anomaly Classification using Artificial Intelligence

The increasing deployment of low-cost industrial IoT (IIoT) sensor platforms on industrial assets enables great opportunities for anomaly classification in industrial plants. The performance of such a classification model depends highly on the available training data. Models perform well when the training data comes from the same machine. However, as soon as the machine is changed, repaired, or put into operation in a different environment, the prediction often fails. For this reason, we investigate whether it is feasible to have a robust and transferable method for AI based anomaly classification using different models and pre-processing steps on centrifugal pumps which are dismantled and put back into operation in the same as well as in different environments. Further, we investigate the model performance on different pumps from the same type compared to those from the training data.

cs.LG↗
arXiv2020

Industrial Federated Learning -- Requirements and System Design

Federated Learning (FL) is a very promising approach for improving decentralized Machine Learning (ML) models by exchanging knowledge between participating clients without revealing private data. Nevertheless, FL is still not tailored to the industrial context as strong data similarity is assumed for all FL tasks. This is rarely the case in industrial machine data with variations in machine type, operational- and environmental conditions. Therefore, we introduce an Industrial Federated Learning (IFL) system supporting knowledge exchange in continuously evaluated and updated FL cohorts of learning tasks with sufficient data similarity. This enables optimal collaboration of business partners in common ML problems, prevents negative knowledge transfer, and ensures resource optimization of involved edge devices.

cs.AI↗
arXiv2019

Monolithically Integrated Perovskite Semiconductor Lasers on Silicon Photonic Chips by Scalable Top-Down Fabrication

Metal-halide perovskites are promising lasing materials for realization of monolithically integrated laser sources, the key components of silicon photonic integrated circuits (PICs). Perovskites can be deposited from solution and require only low temperature processing leading to significant cost reduction and enabling new PIC architectures compared to state-of-the-art lasers realized through costly and inefficient hybrid integration of III-V semiconductors. Until now however, due to the chemical sensitivity of perovskites, no microfabrication process based on optical lithography and therefore on existing semiconductor manufacturing infrastructure has been established. Here, the first methylammonium lead iodide perovskite micro-disc lasers monolithically integrated into silicon nitride PICs by such a top-down process is presented. The lasers show a record low lasing threshold of 4.7 $μ$Jcm$^{-2}$ at room temperature for monolithically integrated lasers, which are CMOS compatible and can be integrated in the back-end-of-line (BEOL) processes.

physics.app-ph↗
arXiv2019

Graphene photodetector integrated on a photonic crystal defect waveguide

We present a graphene photodetector for telecom applications based on a silicon photonic crystal defect waveguide. The photonic structure is used to confine the propagating light in a narrow region in the graphene layer to enhance light-matter interaction. Additionally, it is utilized as split-gate electrode to create a pn-junction in the vicinity of the optical absorption region. The photonic crystal defect waveguide allows for optimal photo-thermoelectric conversion of the occurring temperature profile in graphene into a photovoltage due to additional silicon slabs on both sides of the waveguide, enhancing the device response as compared to a conventional slot waveguide design. A photoresponsivity of 4.7 V/W and a (setup-limited) electrical bandwidth of 18 GHz are achieved. Under a moderate bias of 0.4 V we obtain a photoconductive responsivity of 0.17 A/W.

physics.app-ph↗
arXiv2017

Flexible Hall Sensors Based on Graphene

The excellent electronic and mechanical properties of graphene provide a perfect basis for high performance flexible electronic and sensor devices. Here, we present the fabrication and characterization of flexible graphene based Hall sensors. The Hall sensors are fabricated on 50 um thick flexible Kapton foil using large scale graphene grown by chemical vapor deposition technique on copper foil. Voltage and current normalized sensitivities of up to 0.096 V/VT and 79 V/AT were measured, respectively. These values are comparable to the sensitivity of rigid silicon based Hall sensors and are the highest values reported so far for any flexible Hall sensor devices. The sensitivity of the Hall sensor shows no degradation after being bent to a minimum radius of 4 mm, which corresponds to a tensile strain of 0.6%, and after 1,000 bending cycles to a radius of 5 mm.

physics.app-ph↗
arXiv2017

Graphene photodetectors with a bandwidth larger than 76 GHz fabricated in a 6 inch wafer process line

In recent years, the data traffic has grown exponentially and the forecasts indicate a huge market that could be addressed by communication infrastructure and service providers. However, the processing capacity, space, and energy consumption of the available technology is a serious bottleneck for the exploitation of these markets. Chip-integrated optical communication systems hold the promise of significantly improving these issues related to the current technology. At the moment, the answer to the question which material is best suited for ultrafast chip integrated communication systems is still open. In this manuscript we report on ultrafast graphene photodetectors with a bandwidth of more than 76 GHz well suitable for communication links faster than 100 GBit/s per channel. We extract an upper value of 7.2 ps for the timescale in which the bolometric photoresponse in graphene is generated. The photodetectors were fabricated on 6 inch silicon-on-insulator wafers in a semiconductor pilot line, demonstrating the scalable fabrication of high-performance graphene based devices.

cond-mat.mes-hall↗
arXiv2016

Controlled generation of a pn-junction in a waveguide integrated graphene photodetector

With its electrically tunable light absorption and ultrafast photoresponse, graphene is a promising candidate for high-speed chip-integrated photonics. The generation mechanisms of photosignals in graphene photodetectors have been studied extensively in the past years. However, the knowledge about efficient light conversion at graphene pn-junctions has not yet been translated into high-performance devices. Here, we present a graphene photodetector integrated on a silicon slot-waveguide, acting as a dual-gate to create a pn-junction in the optical absorption region of the device. While at zero bias the photo-thermoelectric effect is the dominant conversion process, an additional photoconductive contribution is identified in a biased configuration. Extrinsic responsivities of 35 mA/W, or 3.5 V/W, at zero bias and 76 mA/W at 300 mV bias voltage are achieved. The device exhibits a 3 dB-bandwidth of 65 GHz, which is the highest value reported for a graphene-based photodetector.

cond-mat.mes-hall↗
arXiv2015

Experimental verification of electro-refractive phase modulation in graphene

Graphene has been considered as a promising material for opto-electronic devices, because of its tunable and wideband optical properties. In this work, we demonstrate electro-refractive phase modulation in graphene at wavelengths from 1530 to 1570 nm. By integrating a gated graphene layer in a silicon-waveguide based Mach-Zehnder interferometer, the key parameters of a phase modulator like change in effective refractive index, insertion loss and absorption change are extracted. These experimentally obtained values are well reproduced by simulations and design guidelines are provided to make graphene devices competitive to contemporary silicon based phase modulators for on-chip applications.

cond-mat.mes-hall↗
arXiv2014

Energy and Performance-Can a Wimpy-Node Cluster Challenge a Brawny Server?

Traditional DBMS servers are usually over-provisioned for most of their daily workloads and, because they do not show good energy proportionality, waste a lot of energy while underutilized. A cluster of small (wimpy) servers, where the number of nodes can dynamically adjust to the current workload, might offer better energy characteristics for these workloads. Yet, clusters suffer from "friction losses" and may not be able to quickly adapt to the workload, whereas a single, brawny server delivers performance instantaneously. In this paper, we compare a small cluster of lightweight nodes to a single server in terms of performance and energy efficiency. We run several benchmarks, consisting of OLTP and OLAP queries at variable utilization to test the system's ability to adjust to the workloads. To quantify possible energy saving and its conceivable drawback on query runtime, we evaluate our implementation on a cluster as well as on a single, brawny server and compare the results w.r.t. performance and energy consumption. Our findings confirm that - based on the workload - energy can be saved without sacrificing too much performance.

cs.DB↗
arXiv2014

Dynamic Physiological Partitioning on a Shared-nothing Database Cluster

Traditional DBMS servers are usually over-provisioned for most of their daily workloads and, because they do not show good-enough energy proportionality, waste a lot of energy while underutilized. A cluster of small (wimpy) servers, where its size can be dynamically adjusted to the current workload, offers better energy characteristics for these workloads. Yet, data migration, necessary to balance utilization among the nodes, is a non-trivial and time-consuming task that may consume the energy saved. For this reason, a sophisticated and easy to adjust partitioning scheme fostering dynamic reorganization is needed. In this paper, we adapt a technique originally created for SMP systems, called physiological partitioning, to distribute data among nodes, that allows to easily repartition data without interrupting transactions. We dynamically partition DB tables based on the nodes' utilization and given energy constraints and compare our approach with physical partitioning and logical partitioning methods. To quantify possible energy saving and its conceivable drawback on query runtimes, we evaluate our implementation on an experimental cluster and compare the results w.r.t. performance and energy consumption. Depending on the workload, we can substantially save energy without sacrificing too much performance.

cs.DB↗