Tecnomatix Plant Simulation Worldwide User Conference 2015 Siemens Industry Software Prof. Dr.-Ing. J. Deuse Plant Simulation Worldwide User Conference 2015 Pushing the Limits of Lean Thinking by Simulation and Data Analytics Univ.-Prof. Dr.-Ing. Jochen Deuse Stuttgart, 16 June 2015 1 dortmund university of technology Prof. Dr.-Ing. J. Deuse Plant Simulation Worldwide User Conference 2015 2 Institute of Production Systems Research Areas Work System Design Time and Motion Studies Digital Manufacturing Manufacturing Data Analytics Human-Robot- Collaboration Factory Physics
12
Embed
Pushing the Limits of Lean Thinking by Simulation and Data ...€¦ · Pushing the Limits of Lean Thinking by Simulation and Data Analytics ... Analytics Human-Robot- ... Optimise
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Tecnomatix Plant Simulation Worldwide User Conference 2015
Siemens Industry Software
Prof. Dr.-Ing. J. Deuse Plant Simulation Worldwide User Conference 2015
Pushing the Limits of Lean Thinkingby Simulation and Data Analytics
Univ.-Prof. Dr.-Ing. Jochen DeuseStuttgart, 16 June 2015
1
dortmunduniversity of technology
Prof. Dr.-Ing. J. Deuse Plant Simulation Worldwide User Conference 2015 2
Institute of Production SystemsResearch Areas
Work System Design
Time and Motion Studies Digital Manufacturing Manufacturing DataAnalytics
Human-Robot-Collaboration
Factory Physics
Tecnomatix Plant Simulation Worldwide User Conference 2015
Siemens Industry Software
Prof. Dr.-Ing. J. Deuse Plant Simulation Worldwide User Conference 2015
Decoding the DNA of theToyota Production System
Basic rules of „Lean Thinking“ (Toyota-DNA)
3
Rule 1: All work shall be highly specified as tocontent, sequence, timing, and outcome.
Stabilise
Rule 2: Every customer-supplier connection mustbe direct, and there must be anunambiguous yes-or-no way to sendrequests and receive responses.
Pull
Rule 3: The pathway for every product and servicemust be simple and direct.
One-Piece-Flow
Rule 4: Any improvement must be made inaccordance with the scientific method,under the guidance of a teacher, at thelowest possible level in the organisation.
Single-Factor Experiments[Spear & Bowen 1999]
Prof. Dr.-Ing. J. Deuse Plant Simulation Worldwide User Conference 2015
Basic rules of „Lean Thinking“ (Toyota-DNA)
4
Rules are formulated generically andtherefore they suggest universaltransferability.
However, the rules must be interpreted in adomain-specific context.Context is defined by Value AddedVariability (VAV).Methods have to be chosen depending onfeatures and expressions of variability, butalso principles have to be scrutinised.
Decoding the DNA of theToyota Production System
Tecnomatix Plant Simulation Worldwide User Conference 2015
Siemens Industry Software
Prof. Dr.-Ing. J. Deuse Plant Simulation Worldwide User Conference 2015
Categories of Variability
5
Does not contribute toadded value
Loss of availabilityLoss of qualityLoss of performance
07 (PZ): Ø=0,6 h, =0,3 h 44 (PZ): Ø=0,9 h, =0,4 hLZ = 50 hLZ = 69 h
VLZ = 1,38
PZ = 35 minPZ = 17 min
VPZ = 0,49
Quantify Variability within Value Streams
Simulate Value Streams using Plant Simulation
Prof. Dr.-Ing. J. Deuse Plant Simulation Worldwide User Conference 2015
Simple/ Obvious: low VAVApply existing methods and tools
sense – categorise – respond
Complex: high VAVExperimental/ Explorative
probe – sense – respond
Classification of Production Systems acc. to Cynefin Framework
14
Chaotic
act – sense – respond
Complicated: high VAVAnalytical/ Experimental
sense – analyse – respond
[acc. to Snowden 2007]
OrderedStatic cause-effect relationship
UnorderedDynamic cause-effect relationship
Tecnomatix Plant Simulation Worldwide User Conference 2015
Siemens Industry Software
Prof. Dr.-Ing. J. Deuse Plant Simulation Worldwide User Conference 2015
Use case: Manufacturing of steam turbine components
15
Analytical/ Experimental:Apply “Factory Physics” and conductsimulation experiments; keep “TrueNorth” of Lean Thinking in mind
Complicated Production Systems
Ordered, complicated systemCause and effect typicallytemporally and spatially distinctDifferent products with differentprocesses and sequences ofoperationsMake-to-Order setting,i. e. unpredictable order sequence
Categories of VariabilityNon VAV:
Loss of availability, quality orperformanceEnvironmental impacts
VAV:Sequence of operationsProcess routingsWork contentDemandLead time
Prof. Dr.-Ing. J. Deuse Plant Simulation Worldwide User Conference 2015
disj
unct
ive
One-Piece-Flow: Layout Housing Manufacturing
16
Optimise factory layout according to Lean Thinking:“The pathway for every product and service must be simple and direct.”
Using simulation experiments to identify capacity overload (Muri)Factory Layout and Flow of Material
pool
ing
ofbo
ttlen
eck
mac
hine
s
Simulation Experiments
Factory Layout and Flow of Material Inventory Utilisation
Tecnomatix Plant Simulation Worldwide User Conference 2015
Siemens Industry Software
Prof. Dr.-Ing. J. Deuse Plant Simulation Worldwide User Conference 2015
Factory Physics: Experiments on Kingman’s Formula
17
Prof. Dr.-Ing. J. Deuse Plant Simulation Worldwide User Conference 2015
Simple/ Obvious: Low VAVApply existing methods and tools
sense – categorise – respond
Complex: High VAVExperimental/ Explorative
probe – sense – respond
Classification of Production Systems acc. to Cynefin Framework
18
Chaotic
act – sense – respond
Complicated: High VAVAnalytical/ Experimental
sense – analyse – respond
[acc. to Snowden 2007]
OrderedStatic cause-effect relationship
UnorderedDynamic cause-effect relationship
Tecnomatix Plant Simulation Worldwide User Conference 2015
Siemens Industry Software
Prof. Dr.-Ing. J. Deuse Plant Simulation Worldwide User Conference 2015
Complex Production Systems
19
Use case: Machining of engine components in a flexible manufacturing system
[Felsomat 2014]
Unordered, complex SystemCausalities are dynamic, butdiscoverableGrowing number of systemelements, interfaces and system-internal relationshipsHigh level of automation
Categories of VariabilityNon VAV:
Varying system loadDynamic bottlenecksLoss of availability, quality orperformance
VAV:Sequence of operationsProcess routingsWork contentDemandLead time
Experimental/ Explorative:Manufacturing Data Analytics basedon unsupervised and supervisedmachine learning
Prof. Dr.-Ing. J. Deuse Plant Simulation Worldwide User Conference 2015
Limited applicability of “Scientific Method”
20
Apply probing instead of Single-Factor ExperimentsIdentify unknown, multivariate patterns in industrial databasesUse data mining, i.e. unsupervised learning procedures
[Felsomat 2014]
System State:
StableUnstable
[RIF e. V.]
Tecnomatix Plant Simulation Worldwide User Conference 2015
Siemens Industry Software
Prof. Dr.-Ing. J. Deuse Plant Simulation Worldwide User Conference 2015 21
Trends in Big Data Analytics
[acc. to Schwab and Keil 2012]
Real-TimeAnalysis
Predictive Data Analysis
Real-TimeReport
ProcessAutomation
UnstructuredExternal Data
Unstructured Internal Data
Traditional InstructionsStrategicPlanning
Zetta(1021)
Exa(1018)
Peta(1015)
Tera(1012)
Years
Data Volume(in bytes)
Ad hoc decisionSupport
OperationalPlanning
AnalysisTime Line
Hig
hD
imen
sion
alD
ata
Seconds Years
Due to long run-times, simulation experiments are not applicable in real-time and predictive analysisSupervised machine learning supports real-time analysisLearning of prediction model by using simulation data
Prof. Dr.-Ing. J. Deuse Plant Simulation Worldwide User Conference 2015
ChaoticNo manageable patternsNo occurrence in production systems
Complicated High VAVApply “Factory Physics” and conductsimulation experimentsKeep “True North” of Lean Thinking in mind
Complex High VAVProbing: unsupervised machine learningCombination of simulation and supervisedmachine learning
Simple/ Obvious Low VAVSelect and apply existing methods and tools
Conclusion
22
AP/Carmelo Imbesi [Kurier.at 2014]
[Felsomat 2014]Level of
IT-Support
Tecnomatix Plant Simulation Worldwide User Conference 2015
Siemens Industry Software
Prof. Dr.-Ing. J. Deuse Plant Simulation Worldwide User Conference 2015
Prof. Dr.-Ing. Jochen Deuse
TU Dortmund UniversityInstitute of Production Systems