Use of data in manufacturing
- 1. © CGI Group Inc. Internal use only. Use of data in Manufacturing ”Sprøjtestøbesektionens Forårsmøde 2018” Jens Christian Volhøj Director, Consulting Services CGI NEXT & Emerging Technology
- 2. Local experts Vertical experience – e.g.: Government, Health, Banking, Insurance, Retail, Utility, Communications, Manufacturing, Logistic, Oil & Gas Leading end-to-end IT and business process services: Business and IT consulting Systems integration IT managed services Business process services 577employees in 5 cities in Denmark Technology expertise – e.g.: Microsoft, Oracle, SAP, Open Source – and more than 150 own developed solutions Strong foundation to support your business goals: Accountability, commitment and more than 40 years of experience Kolding Aarhus Aalborg Ballerup København CGI is the worlds 5th biggest IT service provider First class Business and IT consulting 71.000 employees; approx.. 80% are shareholders Support over 5.000 customers Globally from hundreds of Offices End-to-end- services in IT and Business processes
- 3. Self Service Business Intelligence Big Data Augmented Reality / Virtual Reality Robotic Process Automation / Compliance /GDPR MDM/Data Quality/ Customer journey Industry 4.0Internet of Things (IoT) Dashboards/KPIs Video Analytics Data Warehouse / Data integration Advanced Analytics / Artificial Intelligence Cloud CGI NEXT & Emerging Technology - Focus Areas
- 4. 1778 1882 1969 NOW INDUSTRY 1.0 Mechanization, Steam Power INDUSTRY 2.0 Electricity, Mass Production INDUSTRY 3.0 Automation, Computers, Electronics INDUSTRY 4.0 IoT, Cloud and Cognitive Computing Industrial Revolution
- 5. 5 Digital Strategy IoT - IIoT Advanced Analytics Big Data & Cloud Augmented Reality Cyber Security Application Development Automation “We create value for Industrial customers through innovative data driven solutions for Industry 4.0” Software and IT related areas in Industry 4.0
- 6. Enabling and consolidating data sources 6 Reference data from ERP Files from production equipment (PLC’s with UPC-UA interface) Manufacturing Execution System CSV and Excel files from users Databases Web services IoT sensors Cloud enables powerful data processing and flexible storage
- 7. Data driven value creation in manufacturing with Analytics 7 Optimize Asset Availability & Life Reduce Failures, Optimize Performance Decrease Planned & Unplanned Maintenance Improve Safety Automated Inspection Optimize Workforce Productivity Lower Risk Exposure Optimize Labor & Operation Costs Improve Energy Cost Efficiency Reduce required Compliance Activity Improve forecast Improve Quality
- 8. Video analytics for defect identification • Deep learning artificial intelligence with unsupervised learning within video analytics • Time-efficient and accurate quality checking on every item on the production line • Detect even unforeseen and rare defects • Addition of x-ray imaging can detect defects such as voids in injection molded plastic items, identifying weaknesses difficult to see from the outside • Automatically collect statistics on different defect types to help improve processes sinkhole 0.93 pen 0.98
- 9. Data from plastic molding machines? 9 Examples of data: • Heater Temperature and pressure • Mold temperature and pressure • Mold cooling temperature • Screw rotation speed • Backpressure • Colour concentration and material degradation (Infrared) • Vibrations and sound • Energy consumption • Machine configuration parameters • Product quality and quality issues • Batch data - Mould ID, work shift, operator, materials, etc. • Video or pictures for analysis Examples of Benefits: • Intelligent recommendation for optimal configuration – maybe automated • Reduction of set up time • Best practice “knowledge” data storage • Increased productivity, process optimization and control • Higher OEE (e.g. OEE = Availability × Performance × Quality) • Improved Quality • Scrap reduction • Energy reduction • Improved Sustainability • Shorten time-to-market for new and highly competitive products • Enable smaller batch size • Automated inspection with e.g. Video analytics • Predictive maintenance and less unplanned stops
- 10. Data Driven Implement and monitor the right asset maintenance strategy to meet business drivers and challenges Maintenance Reactive / unplanned Corrective Emergency Proactive / planned Predictive Reliability-centred Condition-based Preventive Constant interval Age based 10 Source: Typology of condition based maintenance – Veldman, Wortmann & Klingenberg
- 11. Enablers for proof of value 11 IoT Explorer Kit Proof of Value use cases InnoLab Training and work with own data Advanced Analytics – Workshop Enabling existing tools like MATLAB Cloud enabling Analytics Fast IoT deployment with pre- defined templates IoT Rapid deployment 1-2 weeks PoV exploration in data set Data Scientist Exploration
- 13. Innovate Use cases Explore Potential Capture Value Sustain Benefits • Innovate • Identify • Prioritise • Analyse • Prepare • Explore • Scale • Implement • Capture value • Maintain • Support • Evolve Approach to data driven value creation
- 14. 14 Commercial estimate of phases Example from customer case Duration: ≈180 hours Output: • Catalogue of prioritized use cases • PoC selection matrix • Recommendations for next steps Estimate: ≈ XXX.000 DKK Duration: ≈ 300 hours Output: • PoC findings and evaluation • Qualification of value creation potential • Recommendations for next steps Estimate: ≈ XXX.000 DKK
- 15. 15 Recommendations • Start with existing data and a known problem to solve - keep it simple. • If data don´t exist then start with simple tools to collect data. Verify your hypnosis before you do large investments. • Find an experienced Data Scientist from similar projects. Fast understanding of your challenges gives fast results. • Get input from external experts this will enhance your results and help keep internal focus on your daily operations. • Create a good foundation for future analytic projects. Share knowledge among multiple employees and document findings. • Use known methodology like the CRISP- model (Cross Industry Standard Process for data mining).
- 16. THANK YOU 16 Jens Christian Volhøj Director, Consulting Services jenschristian.volhoj@cgi.com Lyngbyvej 28, 2100-Copenhagen +45 29 49 89 22