Presentatie SAPIENCE Paper

Report
The control of today and prediction of the
future using predictive production
scheduling
Your logo
Wim De Bruyn
lecturer ICT, R&D FBO, UCG
product mgt. Inxites
Bert Van Vreckem
Lecturer ICT, FBO, UCG
Researcher Prinsyslab, UCG
SAPience.be User Day 2012
1
Agenda
•
•
•
•
•
•
•
•
•
•
Your logo
Introduction
ISA-95: current and future production management
Production Scheduling
ISA-95 and Scheduling: Are they a lovely couple?
ISA-95+Scheduling+SAP = ?
Optimal Schedule and Algorithms
The power of Algorithms on a Real Production Case
From Algorithms to Management View: KPI dashboard
Integration from Shopfloor to Topfloor
Conclusions
SAPience.be User Day 2012
2
Introduction
Your logo
Predictive =
What if we try .....
SAPience.be User Day 2012
3
ISA-95 Functional enterprise
control model
SAPience.be User Day 2012
Your logo
4
ISA-95 Production Information
Overlap
SAPience.be User Day 2012
Your logo
5
ISA-95 Production Segment
Capabilities
SAPience.be User Day 2012
Your logo
6
ISA-95 Product Definition
Model
SAPience.be User Day 2012
Your logo
7
ISA-95 Process segment relations
Connecting a production request with a routing
SAPience.be User Day 2012
Your logo
8
Current and future
Production capabilities
SAPience.be User Day 2012
Your logo
9
ISA-95 Production Schedule
Model
SAPience.be User Day 2012
Your logo
10
Short time
Planning overview
Your logo
• GANTT chart, connecting resources with production orders
(same colour)
SAPience.be User Day 2012
11
Production Scheduling is …
Your logo
• telling a production facility when to make, with which staff, and
on which equipment.
• allocation of jobs to scarce resources
• a combinatorial optimization problem
• maximize and/or minimize objective(s)
• subject to constraints
SAPience.be User Day 2012
12
Production Scheduling
Your logo
• shorten delivery times
• increase variety in end-products
• shorten production lead times
• increase resource utilization
• improve quality, reduce WIP
• prevent production disturbances (machine breakdowns)
 More products in less time!
 Less cost!
 More profit!
 Lower ecological impact!
SAPience.be User Day 2012
13
Production Scheduling in
Manufacturing Planning Framework
Your logo
• Long range prediction and sales planning
• Facility and resources planning
• Demand management, aggregate and workforce
planning
• Order acceptance and resource loading
• Shop floor scheduling, workforce scheduling
SAPience.be User Day 2012
14
Structure of APS
SAPience.be User Day 2012
Your logo
15
Consumer goods planning
Your logo
SAPience.be User Day 2012
16
Agenda
•
•
•
•
•
•
•
•
•
•
Your logo
Introduction
ISA-95: current and future production management
Production Scheduling
ISA-95 and Scheduling: Are they a lovely couple?
ISA-95+Scheduling+SAP = ?
Optimal Schedule and Algorithms
The power of Algorithms on a Real Production Case
From Algorithms to Management View: KPI dashboard
Integration from Shopfloor to Topfloor
Conclusions
SAPience.be User Day 2012
17
Ok!
ERP-system
Information
Transfer
Information
Transfer
Efficiency
Criteria
Numerical
Data
Production
Process
Description
Information
Transfer
Information
Transfer
Optimization
Objective(s)
Variable
Values
Constraints
Information
Transfer
Database
ISA-88 or ISA-95
Compliant and Complete
 = 180
Good Schedule
Production Schedule
 = 150
Automatic
Parameter
Adjustment
Optimization
Algorithm
Compliant? Complete?
Scheduler/
Automatic
Decision
Maker
Manual
Parameter Adjustment
Constraints
Consistency
Check
subject to
22 March 2012
Your logo
Planner/
Decision
Maker
UCG - INXITES R&D
Inconsistent?

  ≤  → ∅
=0
18
Agenda
•
•
•
•
•
•
•
•
•
•
Your logo
Introduction
ISA-95: current and future production management
Production Scheduling
ISA-95 and Scheduling: Are they a lovely couple?
ISA-95+Scheduling+SAP = ?
Optimal Schedule and Algorithms
The power of Algorithms on a Real Production Case
From Algorithms to Management View: KPI dashboard
Integration from Shopfloor to Topfloor
Conclusions
SAPience.be User Day 2012
19
Perfect Plant implementation
(B2MML to SAP at Polar © 2004 World Batch Forum)
SAPience.be User Day 2012
Your logo
20
Mapping SAP PP-PI, ISA95 Production
Schedule, ISA 88 and the Physical Model
Your logo
B2MML to SAP at Polar © 2004 World Batch Forum
SAPience.be User Day 2012
21
Simplified Schedule Request
and Reponse Example
MES
SAP ME
SAP MII
B2MML Production
Schedule XML
(Request)
SAP BC
Your logo
SAP PP
PI
B2MML Production
Schedule XML
(Response)
The schedule can be refined
and adapted in the MES
execution part
Netweaver interface
Web service
SAPience.be User Day 2012
22
Agenda
•
•
•
•
•
•
•
•
•
•
Your logo
Introduction
ISA-95: current and future production management
Production Scheduling
ISA-95 and Scheduling: Are they a lovely couple?
ISA-95+Scheduling+SAP = ?
Optimal Schedule and Algorithms
The power of Algorithms on a Real Production Case
From Algorithms to Management View: KPI dashboard
Integration from Shopfloor to Topfloor
Conclusions
SAPience.be User Day 2012
23
Multiobjective Scheduling
Your logo
 =  → 
 =  → 
  =  → 
SAPience.be User Day 2012
24
Multiobjective Scheduling
Your logo
subject to constraints
 ≤ 
Capacity
Processing
Times
Man Power
SAPience.be User Day 2012
Idle Times
25
Scheduling models
Your logo
• Many models, suitable for specific production
processes
• Continuous: Flow Shop Scheduling
• Discrete: Open/Job Shop Scheduling
• Batch: complex, several models depending on
characteristics
• Not for production scheduling: project
scheduling, timetabling, ...
SAPience.be User Day 2012
26
Scheduling Algorithms
Your logo
General Techniques:
• Mathematical programming
• linear, non-linear, (mixed) integer programming
• Analytical (Exact) methods (enumeration)
• branch-and-bound, branch-and-cut
• dynamic programming
• constraint satisfaction
• Heuristics and meta-heuristics
• genetic algorithm
• tabu search
• Artificial Intelligence
• Reinforcement Learning
• Hybrid Algorithms
SAPience.be User Day 2012
27
Scheduling Algorithms (cont.)
Your logo
• Decomposition Techniques
• Temporal decomposition (rolling horizon
approach)
• Machine decomposition (Shifting Bottleneck)
• Dantzig-Wolf
• MILP-decomposition
SAPience.be User Day 2012
28
Scheduling Algorithms:
Complexity
Your logo
• Analytical techniques
• = algorithms that guarantee optimal solution
• often infeasible
• too many solutions (“NP-hard”)
• mostly suitable for theoretical study of
scheduling problems
SAPience.be User Day 2012
29
Scheduling Algorithms:
Complexity
Your logo
• “Non-analytical” techniques
• = no guaranteed optimum, but feasible in time
• paradigms
• heuristics: use expert knowledge (“rules of
thumb”) to create good schedules
• meta-heuristics: simulated annealing, tabu
search, genetic algorithms (cfr. local
search)
• artificial intelligence: rule-based, agentbased, expert systems
• hybrid: combination of paradigms
SAPience.be User Day 2012
30
Optimization Techniques:
properties
Your logo
• Quality of Solutions Obtained
(How Close to Optimal?)
• Amount of CPU-Time Needed
(Real-Time on a PC?)
• Ease of Development and Implementation
(How much time needed to code,
test, adjust and modify)
• Implementation costs
(Expensive third-party components required?)
SAPience.be User Day 2012
31
Software Solutions w.r.t.
Optimization Techniques
Your logo
Implementation costs
(Expensive LP-solvers required? Easy to
implement?)
Required solution quality?
(Is an immediate answer required, or are long
calculations allowed? Does customer accept
complex solutions?)
SAPience.be User Day 2012
32
How to reduce # searches?
Your logo
Dispatching
Rules
Value
Objective
Function
Local
Search
Beam Search
Branch and Bound
CPU - Time
SAPience.be User Day 2012
33
Decision Support Systems
Your logo
Important issues in design of DSS:
• Database design and management
• Data collection (e.g. barcoding system)
• Module Design and Interfacing
• GUI Design (Gantt-charts, etc.)
• Design of link between GUI and algorithm library
(data organization before transfer)
• Internal Re-optimization
• External Re-optimization
SAPience.be User Day 2012
34
Agenda
•
•
•
•
•
•
•
•
•
•
Your logo
Introduction
ISA-95: current and future production management
Production Scheduling
ISA-95 and Scheduling: Are they a lovely couple?
ISA-95+Scheduling+SAP = ?
Optimal Schedule and Algorithms
The power of Algorithms on a Real Production Case
From Algorithms to Management View: KPI dashboard
Integration from Shopfloor to Topfloor
Conclusions
SAPience.be User Day 2012
35
Real Case
Chemical Batch Production Process
Your logo
(D.Borodin,
B. Van Vreckem,
W. De Bruyn,
MISTA 2011
Scheduling
Conference
Proceedings)
Big
Seed
Fermentation(2)
Main
Fermentation(5)
Buffer tanks (4)
Task
Recovery (1)
• Optimize the Production Process
Objective
• Minimize Total Tardiness (Customer Due-Dates)
Solution Approach
• Exact Optimal Solutions vs. Two Heuristic
Methods
SAPience.be User Day 2012
36
Real Case
Results Comparison
Problem
Instance
Exact
Solution
Your logo
KL
best
KL
time
GA
best
GA
time
N10_1
90
91
7
93
5
N10_2
30
35
16
30
17
N10_3
42
56
10
44
6
N10_4
49
52
14
50
5
N10_5
43
48
6
45
12
N15_1
73
77
80
76
25
N15_2
43
45
112
45
34
N15_3
57
70
39
66
75
N20_1
52
54
490
54
180
N20_2
58
66
260
64
194
N30_1
_
180
1304
186
1560
SAPience.be User Day 2012
37
Agenda
•
•
•
•
•
•
•
•
•
•
Your logo
Introduction
ISA-95: current and future production management
Production Scheduling
ISA-95 and Scheduling: Are they a lovely couple?
ISA-95+Scheduling+SAP = ?
Optimal Schedule and Algorithms
The power of Algorithms on a Real Production Case
From Algorithms to Management View: KPI dashboard
Integration from Shopfloor to Topfloor
Conclusions
SAPience.be User Day 2012
38
GUI’S should allow:
Your logo
• Interactive Optimization
• Freezing Jobs and Re-optimizing
• Creating New Schedules by Combining Different
Parts from Different Schedules
• Cascading and Propagation Effects
After a Change or Mutation by the User, the system:
• does Feasibility Analysis
• takes care of Cascading and Propagation Effects,
• does Internal Re-optimization
SAPience.be User Day 2012
39
GUI for Production Scheduling
•
•
•
•
Your logo
Gantt Chart Interface
Dispatch List Interface
Time Buckets (resource capacity loading)
KPI dashboards
SAPience.be User Day 2012
40
GUI: KPI dashboard
Your logo
Dashboard provides at-a-glance views of key performance indicators (KPIs) relevant to a
particular objective, production or business process: capacities load, costs, profit,
ecological impact, sales, marketing, human resources, etc.
SAPience.be User Day 2012
41
GUI
Important Objectives = KPIs?
Your logo
• Yes and No!
• Due Dates (KPI or objective?)
• Late orders
• Maximum lateness
• Average lateness, tardiness, earliness-tardiness,
makespan
• Productivity and Inventory Related (KPI or objective?)
• Total Setup Time
• Total Machine Idle Time
• Resource usage (KPI or objective?)
• Resource Shortage
SAPience.be User Day 2012
42
Overall KPI concept
Your logo
• KPI-driven Production
• Operations Research Approach:
KPI-driven factory = KPIs as objective functions
• Overall KPI:
 =  +  −  + ⋯ − 
→ min   
where , , … are various KPIs, +  − mean
maximization and minimization of a certain KPI
• Goal: achieve production predictability, lean
manufacturing
SAPience.be User Day 2012
43
Agenda
•
•
•
•
•
•
•
•
•
•
Your logo
Introduction
ISA-95: current and future production management
Production Scheduling
ISA-95 and Scheduling: Are they a lovely couple?
ISA-95+Scheduling+SAP = ?
Optimal Schedule and Algorithms
The power of Algorithms on a Real Production Case
From Algorithms to Management View: KPI dashboard
Integration from Shopfloor to Topfloor
Conclusions
SAPience.be User Day 2012
44
Integration from shopfloor
to topfloor
Your logo
V.P.
Mfg
Mgr.
Mgr.
Mgr.
SAP MII
DCS / PLC
MES
Plant Historian
SCADA / HMI
LIMS/ Inspection /
Equipment Testing
Plant DB
Environmental
Building Management
Plant Data
Collection
SAPience.be User Day 2012
Wireless Integration
45
SAP MII as KPI dashboard and
and production schedule
interface
SAP MII extracts data
from SAP ERP and
provides real-time
visibility and
distribution to Plant
Floor Systems of:
SCM
ERP
Your logo
SAP MII’s ability to perform
transaction execution into
SAP also enables
automated, plant-level
creation of:
PLM
Global Coordination
PLAN
MAKE
DELIVER
Manufacturing Integration & Intelligence
Business Logic Services
•Planned Orders
•Bills of Material
•Production & Process
Orders
•Material Inventory Levels
•Inspection Lots Data
•Master Recipes
•Material Details
•Batch Details
•Resources & Functional
Locations
•Maintenance Work Order
& Notification details
•Material & Order Costs
Visualization
Composition Environment
Quality Engine
KPIs / Metrics / Alerts
Data Services
Simplified Execution
Plant Historian
DCS / PLC via
OPC
MES
22 March 2012
•Production Confirmations
•Process Messages
•Material Receipts
•Material Consumptions
•Material transfers
•Inspection results recording
•Quality Notifications
•Batch Characteristic recording
•Work Orders & results recording
•Maintenance Notifications
LIMS/ Inspection /
Equipment Testing
MAKE
SAPience.be User Day 2012
Plant DB
Environmental
Building Management
SCADA / HMI
Wireless Integration
46
Better Asset usage, right product at the
right time, less inventory, less waste
“
Your logo
Reduce throughput times 30%
Reduce inventory 15 – 20%
VP Mfg
14% less errors in production
Quartile 1: 95% OEE vs. 78% Average*
Reduce data capture efforts 65%
Source: MESA International
!
Plant Mgr.
Machine Uptime 99.5% Qtr 1
vs. 93.6% Average*
Qtr. 1: 98% First pass quality vs. 75%
Avg.*
* Benchmarks from ASUG Manufacturing Benchmarking Study
Qtr. 1: 98.5% On-Time Delivery vs.
89.1% Avg.*
SAPience.be User Day 2012
47
Interaction between SAP APO
and Workcenter schedule
Your logo
Interactive
Workcenter
Schedule: Production
Schedule updated
every 30 minutes.
Double Clicking on a
production order provide
the confirmation screen
to enter new production.
SAPience.be User Day 2012
48
More control – more interactivity – More
planner and operator responsibility
Alert !
Dashboards
für die
intelligente
Fertigung
SAP BI
Weitere SAP-Lösungen
SAP NETWEAVER
Your logo
 Invoking of scheduling solution
(SAP SCM APO).
 Machine disruption is considered
as machine-breakdown in
scheduling board.
 Finding an alternative capacity
(manually or through rescheduling
run).
SAP Manufacturing
(mySAP ERP)
SAP MII
Manufacturing Intelligence
Manufacturing Integration
Planner is able to respond to
disruption in realtime and to resolve
the conflict.
SAPience.be User Day 2012
49
Agenda
•
•
•
•
•
•
•
•
•
•
Your logo
Introduction
ISA-95: current and future production management
Production Scheduling
ISA-95 and Scheduling: Are they a lovely couple?
ISA-95+Scheduling+SAP = ?
Optimal Schedule and Algorithms
The power of Algorithms on a Real Production Case
From Algorithms to Management View: KPI dashboard
Integration from Shopfloor to Topfloor
Conclusions
SAPience.be User Day 2012
50
Conclusions
Your logo
Optimization of Production Scheduling (Algorithms)
implemented in the Standardized Environment (ISA-95compliant) and Incorporated in the Production Automation
System (SAP) that allows visibility and transparency for all
stakeholders involved in Production Process (KPI dashboard),
will enable to:
Reduce Cost
Increase
Profit
Respect the
Environment
= Optimize
Ecological
Footprint
SAPience.be User Day 2012
Realise Lean
Production:
avoid waste
and timeloss
51
Thank you!
Your logo
SAPience.be User Day 2012
52

similar documents