Publications

2023

1. Automation of Experimental Modal Analysis Using Bayesian Optimization

   Johannes Ellinger, Leopold Beck, Maximilian Benker, and 2 more authors

   Applied Sciences 2023

   Abs HTML

   The dynamic characterization of structures by means of modal parameters offers many valuable insights into the vibrational behavior of these structures. However, modal parameter estimation has traditionally required expert knowledge and cumbersome manual effort such as, for example, the selection of poles from a stabilization diagram. Automated approaches which replace the user inputs with a set of rules depending on the input data set have been developed to address this shortcoming. This paper presents an alternative approach based on Bayesian optimization. This way, the possible solution space for the modal parameter estimation is kept as widely open as possible while ensuring a high accuracy of the final modal model. The proposed approach was validated on both a synthetic test data set and experimental modal analysis data of a machine tool. Furthermore, it was benchmarked against a similar tool from a well-known numerical computation software application.

2. Exposure time and point cloud quality prediction for active 3D imaging sensors using Gaussian process regression

   Alejandro Magaña, Lukas Schneider, Maximilian Benker, and 3 more authors

   Production Engineering 2023

   Abs HTML

   Setting an optimal image exposure is crucial for acquiring dense point clouds using 3D active optical sensor systems such as structured light sensors [structured light sensors (SLSs)] and active stereo sensors. One of the most common and seamless ways to optimize the image brightness of an image exposure is to adjust the camera’s exposure time. However, optimizing the image exposure alone is ineffective for acquiring surfaces of large-scale objects with a complex topology if a spatial understanding of the scene is neglected. Hence, the present paper proposes a data-driven approach using two Gaussian processes [Gaussian processes (GPs)] regression models to select a proper exposure time considering the nonlinear correlations between image exposure and the scene spatial relationships. To model these correlations, our study introduces first the generic synthesization of seven inputs and two target variables. Then, based on these inputs, two independent GPs are designed: one for predicting the measurement quality and one for estimating the exposure time. The performance and generalizability of both models are thoroughly evaluated using an SLS and an active stereo sensor. The evaluation demonstrated that the point cloud quality models adequately matched observations with an R2 exceeding 90%. Specifically, the models predicted point cloud quality with an root mean square error (RMSE) of 10%. Additionally, the assessment of the performance of the exposure time models showed a model fit with an R2 above 97%. The exposure time prediction accuracy, as evidenced by the RMSE values, was within 10% of the corresponding exposure time range for each sensor. The present research shows the potential and effectiveness to completely automate the assessment of a point cloud quality and the selection of exposure times with the help of data-driven models.

3. Condition Monitoring of Machine Tool Feed Drives and Methods for the Estimation of Remaining Useful Life

   Maximilian Benker

   2023

   Abs HTML

   Condition-based maintenance of machine tool feed drives offers great potential to increase the efficiency in industrial manufacturing. In this publication-based dissertation, machine learning methods were developed and applied for condition monitoring of machine tool feed drives and the estimation of remaining useful life. The proposed approaches showed high prediction accuracies even in cases where only a few historical observations were available.

2022

1. Condition Monitoring of Ball Screw Feed Drives Using Convolutional Neural Networks

   Maximilian Benker, and Michael F. Zaeh

   CIRP Annals 2022

   Abs HTML

   Ball screw feed drives are widely used in machine tools and significantly determine the manufacturing quality and efficiency. With their degradation, machining accuracy and economic efficiency decrease. Therefore, monitoring the condition of ball screws is of great interest. Past investigations showed that condition monitoring of ball screws is possible. Nevertheless, practical applications of a condition monitoring system for ball screw drives are rare, as it is unclear how well they perform on unseen components. In this paper a data-driven approach is presented, which can assess the condition of unseen ball screws with an accuracy of up to 98%.

2. Experimental Derivation of a Condition Monitoring Test Cycle for Machine Tool Feed Drives

   Maximilian Benker, Sebastian Junker, Johannes Ellinger, and 2 more authors

   Production Engineering 2022

   Abs HTML

   Due to their critical influence on manufacturing accuracy, machine tool feed drives and the monitoring of their condition has been a research field of increasing interest for several years already. Accurate and reliable estimates of the current condition of the machine tool feed drive’s components ball screw drive (BSD) and linear guide shoes (LGSs) are expected to significantly enhance the maintainability of machine tools, which finally leads to economic benefits and smoother production. Therefore, many authors performed extensive experiments with different sensor signals, features and components. Most of those experiments were performed on simplified test benches in order to gain genuine and distinct insights into the correlations between the recorded sensor signals and the investigated fault modes. However, in order to build the bridge between real use cases and scientific findings, those investigations have to be transferred and performed on a more complex test bench, which is close to machine tools in operation. In this paper, a condition monitoring test cycle is developed for such a test bench. The developed test cycle enables the recording of a re-producible data basis, on which models for the condition monitoring of BSDs and LGSs can be based upon.

3. An Investigation into the Economic Efficiency of Different Maintenance Strategies Based on a Discrete Event Simulation

   Maximilian Benker, Victor Rommel, and Michael F. Zaeh

   Procedia CIRP 2022

   Abs HTML

   Many companies remain reluctant towards the implementation of a predictive maintenance (PdM) strategy, as investment costs for that are high *aMCnadonrrtyehsecpoormenmdpinagniniaeiunstghroeurms.eTafieunll.:rl+eifl3ue3tci3tma8ne7t(3tRo7wU5aL4r)3d0es;sttEhim-emaiamtielpsa,ldewdmrheeisncsth:aptiiatouinls.sobtifaesfa@epderneosdnai,mcat.ireveueumncaeinrtaeinna.nHcen(cPed,Mit)istruantcelgeya,rahsoiwnvvesiatmblenat cPodsMts fsotrrathegatyairse. Thihgihs panapdetrhiesraesmseasisniinngg tuhseefeuclolniofemtiimc eim(RpaUcLt )ofesatiPmdaMtess,trwahteigchy iwt iitshbtahseehdeolpn,oafrae duinsccerrettaeine.vHenetnsciem, uitlaitsiounncmleoadrehl,olweavdianbgletoathPedMinssitgrhatetghyatisa. PTdhMis sptarpateergiys oausstpesesrifnogrmthseoethceornsotmraitcegiimespaecvtenofina cPadsMes ostfrahtiegghylywuinthcetrhteaihneRlpUoLfpareddisictrieotnese.vIet nletasdims tuolaatnioincmreoadsel,inletaodtainl grutnontihneg itnimsieghutptthoat5a%PadnMd 4str%at,ecgoymopuatrpeedrftoorma sreoatchteivrestarnadtegaipelsaenvneendimn aciansteesnoanf cheigshtrlayteugnyc,eretasipnecRtUivLelyp.redictions. It leads to an increase in total running time up to 5 % and A4b%st,rcaocmt pared to a reactive and a planned maintenance strategy, respectively.

2021

1. A Gaussian Process Based Method for Data-Efficient Remaining Useful Life Estimation

   Maximilian Benker, Artem Bliznyuk, and Michael F. Zaeh

   IEEE Access 2021

   Abs HTML

   The task of remaining useful life (RUL) estimation is a major challenge within the field of prognostics and health management (PHM). The quality of the RUL estimates determines the economical feasibility of the application of predictive maintenance strategies, that rely on accurate predictions. Hence, many effective methods for RUL estimation have been developed in the recent years. Especially deep learning methods have been among the best performing ones setting new record accuracies on bench mark data sets. However, those approaches often rely on numerous and representative run-to-failure sequences of the components under investigation. In real-world use cases, this kind of data (i.e. run-to-failure sequences and RUL labels) is hardly ever present. Therefore, this paper proposes a new, data-efficient method, which is based on Gaussian process classification to derive abstract health indicator (HI) values in a first step, and warped, monotonic Gaussian process regression for indirect RUL estimation in a second step. The proposed approach does neither rely on entire run-to-failure sequences nor on any RUL labels and was tested on the benchmark C-MAPSS turbo fan and FEMTO bearing data sets, achieving comparable results to the state-of-the art whilst using only a small fraction of the available training data. Hence, the proposed approach allows RUL estimation in use cases, in which gathering enough failure data for the application of deep learning models is infeasible.

2. Utilizing Uncertainty Information in Remaining Useful Life Estimation via Bayesian Neural Networks and Hamiltonian Monte Carlo

   Maximilian Benker, Lukas Furtner, Thomas Semm, and 1 more author

   Journal of Manufacturing Systems 2021

   Abs HTML

   The estimation of remaining useful life (RUL) of machinery is a major task in prognostics and health management (PHM). Recently, prognostic performance has been enhanced significantly due to the applicationof deep learning (DL) models. However, only few authors assess the uncertainty of the applied DL models and therefore can state how certain the model is about the predicted RUL values. This is especially critical in applications, in which unplanned failures lead to high costs or even to human harm. Therefore, the determination of the uncertainty associated with the RUL estimate is important for the applicability of DL models in practice. In this article, Bayesian DL models, that naturally quantify uncertainty, were applied to the task of RUL estimation of simulated turbo fan engines. Inference is carried out via Hamiltonian Monte Carlo (HMC) and variational inference (VI). The experiments show, that the performance of Bayesian DL models is similar and in many cases even beneficial compared to classical DL models. Furthermore, an approach for utilizing the uncertainty information generated by Bayesian DL models is presented. The approach was applied and showed how to further enhance the predictive performance.

2020

1. Predicting the Ultimate Tensile Strength of Friction Stir Welds Using Gaussian Process Regression

   Roman Hartl, Fabian Vieltorf, Maximilian Benker, and 1 more author

   Journal of Manufacturing and Materials Processing 2020

   Abs HTML

   In the work described here, Gaussian process regression was applied to predict the ultimate tensile strength of friction stir welds through data evaluation and to therefore avoid destructive testing. For data generation, a total of 54 welding experiments were conducted in the butt joint configuration using the aluminum alloy EN AW-6082-T6. Four tensile samples were taken from each of the 54 experiments and the resulting ultimate tensile strength of the weld seam segment was modeled as a function of the weld’s surface topography. Further models were created for comparison, which received either the process variables or the process parameters to predict the ultimate tensile strength. It was shown that the ultimate tensile strength can be accurately predicted based on the weld’s surface topography. Especially for low welding speeds, the correlation coefficients between the true and the predicted ultimate tensile strength were high. However, overall, even higher correlation coefficients could be achieved when providing the process variables or the process parameters to the model. In conclusion, it was shown that the developed Gaussian process regression model is a powerful approach for replacing destructive testing and for predicting ultimate tensile strength based solely on data that can be collected non-destructively.

2019

1. Estimating Remaining Useful Life of Machine Tool Ball Screws via Probabilistic Classification

   Maximilian Benker, Robin Kleinwort, and Michael F. Zaeh

   In 2019 IEEE International Conference on Prognostics and Health Management (ICPHM) 2019

   Abs HTML

   Ball screws are key components in machine tool linear feed drives since they translate the motors’ rotary motion into linear motion. With usage over time, however, tribological degradation of ball screws and the successive loss in preload can cause imprecise position accuracy and loss in manufacturing precision. Therefore condition monitoring (CM) of ball screws is important since it enables just in time replacement as well as the prevention of production stoppages and wasted material. This paper proposes an idea based on a probabilistic classification approach to monitor a ball screw’s preload condition with the help of modal parameters identified from vibration signals. It will be shown that by applying probabilistic classification models, uncertainties with respect to degradation can be quantified in an intuitive way and therefore can enhance the basis of decision making. Furthermore, it will be shown how a probabilistic classification approach allows the estimation of remaining useful life (RUL) for ball screws when the user only has access to discrete preload observations.

2. Feed Drive Condition Monitoring Using Modal Parameters

   Johannes Ellinger, Thomas Semm, Maximilian Benker, and 3 more authors

   MM Science Journal 2019

   Abs HTML

   Ball screws and linear guides are among the key components of machine tools. Abrasive wear causes a loss in stiffness of these components over time affecting the attainable manufacturing precision and, eventually, leads to failures and costly down-time. In order to control these effects, the condition of the crucial feed drive components needs to be monitored. This paper shows, how the feed drive condition can be monitored by looking at the modal parameters of the system. It will be shown, that preload loss cannot only be detected globally, but can be traced back to the worn component. A distinct test cycle was developed for this purpose.