Industrial Sensory Data Analytics Introduction, Analysis Goals & - PowerPoint PPT Presentation

Industrial Sensory Data Analytics

Introduction, Analysis Goals & Methods/Tools

Industrial Sensory Data Analytics Introduction Pattern Recognition in Time Series Data ▪ The field of Natural Language Processing (NLP) has seen a number of deep learning based advances. ▪ These examples include speech recognition, voice recognition and speaker separation, and are based on recognizing patterns in the time series data that characterizes sound. Analog to digital Pattern Speaker Analog signal Application conversion recognition Industrial Sensory Data Analytics 3 Chair of Technologies and Management of Digital Transformation, University of Wuppertal

Industrial Sensory Data Analytics Introduction Pattern Recognition in Time Series Data ▪ The field of Natural Language Processing (NLP) has seen a number of deep learning based advances. ▪ These examples include speech recognition, voice recognition and speaker separation, and are based on recognizing patterns in the time series data that characterizes sound. Analog to digital Pattern Tool Analog signal Quality conversion recognition Industrial Sensory Data Analytics 4 Chair of Technologies and Management of Digital Transformation, University of Wuppertal

Industrial Sensory Data Analytics Analysis Goals Signal Classification & Anomaly Detection Signal Forecasting Soft Sensors & Signal Construction Interpretability and Explainability Industrial Sensory Data Analytics 5 Chair of Technologies and Management of Digital Transformation, University of Wuppertal

Industrial Sensory Data Analytics Analysis Methods and Tools ▪ Supervised Machine Learning for Classification and Anomaly Detection. ▪ Unsupervised Machine Learning and Dynamic Time Warping for Similarity Analysis. ▪ Deep Learning and Long-Short-Term Memory Networks for Time Series Forecasting. ▪ Web Based Development Tools for Interactive Visualization Dashboards. Industrial Sensory Data Analytics 6 Chair of Technologies and Management of Digital Transformation, University of Wuppertal

Industrial Applications Selected Research Projects and Applications

Industrial Applications Deep Drawing of Car Body Parts Selected Research Projects and Applications

Deep Drawing of Car Body Parts The Product ▪ Manufacturing of car body parts with a deep drawing tool. ▪ ~80 different parts with variable size, shape and curvature. in collaboration with Industrial Sensory Data Analytics 9 Chair of Technologies and Management of Digital Transformation, University of Wuppertal

Deep Drawing of Car Body Parts The Manufacturing Process Coarse cutting Die cutting Deep drawing Water jet cutting Cleaning Quality Control Industrial Sensory Data Analytics 10 Chair of Technologies and Management of Digital Transformation, University of Wuppertal

Deep Drawing of Car Body Parts The Deep Drawing Tool Extension of the tool by a sensor system for continuous data acquisition during manufacturing. Using the sensory data for process monitoring and failure prediction . Industrial Sensory Data Analytics 11 Chair of Technologies and Management of Digital Transformation, University of Wuppertal

Deep Drawing of Car Body Parts The Data ▪ Strain gauge sensors give information about the force exerted upon the metal sheet. ▪ Course of the sensor signal indicates process failures. Industrial Sensory Data Analytics 12 Chair of Technologies and Management of Digital Transformation, University of Wuppertal

Deep Drawing of Car Body Parts The Data ▪ Strain gauge sensors give information about the force exerted upon the metal sheet. ▪ Course of the sensor signal indicates process failures. Industrial Sensory Data Analytics 13 Chair of Technologies and Management of Digital Transformation, University of Wuppertal

Deep Drawing of Car Body Parts The Learning Problem ▪ Forecast of the strain gauge signal based on a limited cutout. ▪ Extraction of relevant information to predict the course of the sensor signal. Prediction Network Input Target Regular Production Cracking Metal Sheet Industrial Sensory Data Analytics 14 Chair of Technologies and Management of Digital Transformation, University of Wuppertal

Deep Drawing of Car Body Parts The Learning Model Strain gauge Strain gauge cutout forecast Regressor Classifier bi-LSTM bi-LSTM bi-LSTM bi-LSTM Output LSTM LSTM LSTM Input Input 2 30 30 30 128 128 128 128 Industrial Sensory Data Analytics 15 Chair of Technologies and Management of Digital Transformation, University of Wuppertal

Deep Drawing of Car Body Parts Results of the Signal Forecast Industrial Sensory Data Analytics 16 Chair of Technologies and Management of Digital Transformation, University of Wuppertal

Deep Drawing of Car Body Parts In-line Signal Forecasting & Failure Prediction From manual labor and visual inspection of the product to automated and algorithmic data driven quality control. Industrial Sensory Data Analytics 17 Chair of Technologies and Management of Digital Transformation, University of Wuppertal

Deep Drawing of Car Body Parts Interpretability of the Model’s Decision What are the important parts of the signal that lead to the prediction of a process failure? Baseline (trained on the complete signal) Prediction case (trained on the cutout signal) Industrial Sensory Data Analytics 18 Chair of Technologies and Management of Digital Transformation, University of Wuppertal

Industrial Applications Soft Sensors for Prototype Vehicles Selected Research Projects and Applications

Soft Sensors for Prototype Vehicles The Product, Task & Goal Determination of operating strengths and design loads for kinematic components in production vehicles. Extension of the car chassis of prototype vehicles with sensors to acquire load data during operation. Development of soft sensors for series production vehicles . in collaboration with Industrial Sensory Data Analytics 20 Chair of Technologies and Management of Digital Transformation, University of Wuppertal

Soft Sensors for Prototype Vehicles The Soft Sensor Training Process Internal control units Preprocessing External sensors Soft sensor models Industrial Sensory Data Analytics 21 Chair of Technologies and Management of Digital Transformation, University of Wuppertal

Soft Sensors for Prototype Vehicles Evaluating Soft Sensors in Test Drives Saving of electronics/hardware in series production vehicles by replacing real sensors with AI driven soft sensor models. Industrial Sensory Data Analytics 22 Chair of Technologies and Management of Digital Transformation, University of Wuppertal

Industrial Applications Failure Classification for Railway Switches Selected Research Projects and Applications

Failure Classification for Railway Switches Problem setting and Analysis Goal 60k switches in German railway network maintained by DB Netz AG. Remote monitoring of switching operations via current signals. Deep learning based classification of current signals to identify faulty operations in collaboration with Industrial Sensory Data Analytics 24 Chair of Technologies and Management of Digital Transformation, University of Wuppertal

Failure Classification for Railway Switches Classification of Characteristic Failure Offsets Pre-processing Failure Classification Trimming 1D-Convolutional Neural Network Input: Database  Operation Time Historical Signals of ~ 5k Switches DTW Matching Input: Global FC Output Convolutions Offset Average Layer (Softmax) Offset Extraction (128, 256, 128) Signal Pooling BatchNorm, ReLU Industrial Sensory Data Analytics 25 Chair of Technologies and Management of Digital Transformation, University of Wuppertal

Failure Classification for Railway Switches Overcoming Sparse Data by Semi-Supervised Training Low accuracy of traditional ML model Deep neural network requires many manually labeled curves (i.e. failures) Data enrichment by using ML-Tool for decision support and labeling Nearest neighbor Validated 1D-convolutional Small data catalog classifier training data neural network Classification Classification accuracy: 84% accuracy: 97% Classifier validation tool Reducing time to maintenance in case of operation failures! Industrial Sensory Data Analytics 26 Chair of Technologies and Management of Digital Transformation, University of Wuppertal

Richard Meyes, M.Sc. Tel: +49 (0)202 439 1046 meyes@uni-wuppertal.de Chair for Technologies and Management of Digital Transformation Univ. Prof. Dr. Ing. Tobias Meisen https://www.tmdt.uni-wuppertal.de/ Campus Freudenberg Rainer-Gruenter-Str. 21 D-42119 Wuppertal Germany University of Wuppertal School of Electrical, Information and Media Engineering