Industrial Big Data and Digitalization in production with Predictive Analytics Big data cloud for data enabling and storage Data from ERP, MES, Factories, Production lines, Machines and QA Industrial IoT for data enabling Close data gaps and add new data streams Grundfos use case om Industry 4.0 Big Data cloud arkitektur for IIoT, predictive analytics og Rigtig Industry 4.0
Introduktion Kim H. Nikolajsen - Business Advisor / Value Architect / Team lead Industrial PROFILE Passionate about business development and digitalization. 20 years of experience with digital strategy, and enterprise software development are used when analyzing business digitalization. From Tech start-ups Kim are experienced with various business models and their disrupting impact on traditional enterprises. Team lead on Grundfos case! MAIN FOCUS Digitalization in Manufacturing Industry 4.0 Internet of Thing (IoT) Big Data Advanced Analysis. Business Intelligence 2
Grundfos - Projects and Goals Business goals Improve quality Reduce waste Optimize processes Minimize product recalls Two explorative projects to get increased knowledge about: Big Data in Production - A Pilot / PoC Data enabling from production systems (MES) and Machines Industrial IoT for closing data gaps Visualization of production data Designing the IT Architecture for Big Data in an Industrial production Test IoT Cloud platform Prediction in production Inspire and Test predictively scenarios in production processes Data sources: MES and test data
Project details
Grundfos pump production Multible data sources from factories, production lines and sub contractors Sub contractors Electronic production Pump housing production Multiple Production lines
Closing data gaps Enabling and consolidating data sources Reference data from ERP Manufactoring Execution System Files from production equipment (PLC s with UPC-UA interface) CSV and Excel files from users Databases Web services IoT sensors
Closing data gaps with IIoT Rapid deployment IoT Kit Sensor use case examples: Air quality in production (temperature, particles, humidity) Employee Proximity machine (security, movement heat map) Industrial IoT Rapid deployment concept 1) Setup IoT Gateway 10 minutes 2) Mounting and connecting 4 sensors 12 minutes 3) Register further metadata for easy administration 8 minutes Windows 10 team images makes it easy for IT to automate updates and deployment
Big Data architecture - overview Streaming data for current information and fast event based action Cold data path for storage, reporting and long term analytics MES Production Line Webservice MES ERP PLC/ Kepware Power BI Lake Analytic
Flow controller production Bill of Material (BOM) and processes Sub contractors and component manufacturing Electronic production Flow controller production line Flow controller for pump controller mounted on pipe for house hold circulator
Predictive Analytics failure in production Will used components cause failure in later test? Custom visualisations in Power BI Data sources used: BOM & Test data Flow controller for pump Predicted result with 89% accuracy Predictive model tested for controlling Manufacturing Execution System (MES) - enabling Industry 4.0
Some visulizations and Dashboards from: Grundfos Big Data in electronic production
Streaming Dash boards (Live data) Dashboard with IoT sensors Temperature
Custom Visuals in Power BI Current live status on machines in production line Green: < 50 Yellow: 51 250 Red: > 250 Total status for the day: Count on shift in status Many shift of the machine status indicate problems or bottlenecks
ARBI App used in production Using Augmented Reality for visualizing data feed from Azure cloud
Conclusion on project benefits Grundfos statement Successful proof of concept with complete buy-in to concept from entire organization, from process owners to executive management Inspirational showcase for future IoT/Big Data projects 89 % accuracy in test failure prediction in production = significant potential cost savings Awareness for Advanced analytics with Proof of Concept Flexible and scalable platform enables pay-as-you-grow approach Strategy and roadmap for next steps and full scale implementation
For further information Kim Nikolajsen Value Architect, Team Manager Industrial +45 61717239 kim.nikolajsen@affecto.com https://linkedin.com/in/kimnikolajsen