Post on 11-Dec-2021
transcript
Intern © Siemens 2019
Digital Twin of Performance
MindSphere @Tübingen
Ist die Digitalisierung in der Cloud nur etwas für Konzerne?
Intern © Siemens 2019
Key Topics
Seite 2 Yunus Arslan
Does collecting production data actually gain value?
Which data is the right data?
How do I use the data for my benefit?
08.10.2020
Intern © Siemens 2019Seite 5 Yunus Arslan
2 // Digital Paint Shop Project
Key facts
• Cost several Mio. €
• Replaces 20-year-old paint shop
• Designed in Siemens NX
• Simulated throughput with Plant Simulator
• MindSphere connection
08.10.2020
Intern © Siemens 2019Seite 6 Yunus Arslan
2 // Digital Paint Shop Project
How high is the throughput /availability?
How high are the running costs?
How can we get less downtime?
How can we improve our quality?
Business understanding
08.10.2020
Intern © Siemens 2019Seite 8 Yunus Arslan
3 // Requirements for the MindSphere Use Cases
Go online on day one!
OEE/KPI ✓ Maintenance /service ✓
Operating costs ✓Quality control ✓
08.10.2020
Intern © Siemens 2019Seite 9 Yunus Arslan
3 // Requirements for the MindSphere Use Cases
Business understanding
Data understanding
Data preperation
Modeling
Evaluation
Deployment
Data
understandingCRISP-DM
How we did it:
• Instructing the producer to the MindSphere structure and
capabilities even before PLC program was created
• Defined over 160 data points collaborative with the producer -
about 50% got defined by the producer itself to benefit from his
product knowledge
Always keep the target in mind while identifying the data points!
Always use the experience of the process or product owner!
08.10.2020
Intern © Siemens 2019Seite 10 Yunus Arslan
3 // Requirements for the MindSphere Use Cases
Data
preperation
08.10.2020
Intern © Siemens 2019Seite 11 Yunus Arslan
Data aggregation
• MindConnect Nano connected to S7 PLC
Data modeling
• VFC to model data• Count Gearboxes • Calculate KPI• Calculate monthly
consumptions
Visualization
• easyDash to Visualize Time Series Data
• New Application • Based on open source
software• Uses MindSphere
Timeseries API
Data modeling
Data preperation
Deployment
3 // Requirements for the MindSphere Use Cases
08.10.2020
Intern © Siemens 2019Seite 13 Yunus Arslan
Performance Analysis
ü Clear and easy to reach data of production throughput
ü Analysis of bottlenecks ü Calculation of interruption costs
Already used to calculate “Corona break” loss.
Service
ü Clear visualization of the plant status
ü Automatic message to responsible persons
ü Early detection of problemsü Faster reaction time
Already detected several problems.
Maintenance
ü Time savings for facility management and maintenance
ü Historical data of plant
Already in use to make expense report.
4 // Achievements
How high is the throughput? How can we get less downtime? How high are the running costs?
08.10.2020