from design of experiments to closed loop control · from design of experiments to closed loop...
TRANSCRIPT
![Page 1: From Design of Experiments to closed loop control · From Design of Experiments to closed loop control Petter Mörée & Erik Johansson Umetrics](https://reader030.vdocuments.us/reader030/viewer/2022041204/5d52c42288c993073e8b712d/html5/thumbnails/1.jpg)
From Design of Experiments to closed loop control
Petter Mörée & Erik Johansson
Umetrics
![Page 2: From Design of Experiments to closed loop control · From Design of Experiments to closed loop control Petter Mörée & Erik Johansson Umetrics](https://reader030.vdocuments.us/reader030/viewer/2022041204/5d52c42288c993073e8b712d/html5/thumbnails/2.jpg)
Umetrics, The Company • Part of ~1Billion conglomerate • The market leader in software for multivariate
analysis (MVDA) & Design of Experiments (DOE)
• 25+ years in the market • Off line analysis tools • On-Line process monitoring and fault detection • 700+ companies, 7,000+ users • Pharmaceutical, Biotech, Chemical, Food,
Semiconductors and more • Worldwide Presence with MKS • Offices:
– Umeå, Malmo, Sweden – York, England – Boston, San Jose, USA – Singapore – Frankfurt, Germany
• Close collaboration with universities in USA, Sweden, UK and Canada
• Partnership with Sartorius; global marketing, distribution, development and integration.
![Page 3: From Design of Experiments to closed loop control · From Design of Experiments to closed loop control Petter Mörée & Erik Johansson Umetrics](https://reader030.vdocuments.us/reader030/viewer/2022041204/5d52c42288c993073e8b712d/html5/thumbnails/3.jpg)
3
Building a capable process
DOE
MVDA
QFD Quality Function Deployment
QRA: Quality Risk Assessment
DOE Analysis Design Space
Control Strategy
Manufacturing
• DOE is a knowledge building tool for process development • MVDA is used both for process understanding and process monitoring
![Page 4: From Design of Experiments to closed loop control · From Design of Experiments to closed loop control Petter Mörée & Erik Johansson Umetrics](https://reader030.vdocuments.us/reader030/viewer/2022041204/5d52c42288c993073e8b712d/html5/thumbnails/4.jpg)
• Multivariate data analysis (MVA) is a tool to learn from data
• Marek used MVA and NIR to predict glucose nad other parameters inside the reactor
• This talk will focus on process parameters – Tightly controlled
• pH, pO2, Temperature – Parameters used for keeping tightly controlled at their sepoint
• Stirring, airation, cooling, base addition ..
– Commonly measured • CER, OUR …
• Monitor, interpret, control
4
Processes and their data are never perfect Delegates at this meeting are of course excluded
![Page 5: From Design of Experiments to closed loop control · From Design of Experiments to closed loop control Petter Mörée & Erik Johansson Umetrics](https://reader030.vdocuments.us/reader030/viewer/2022041204/5d52c42288c993073e8b712d/html5/thumbnails/5.jpg)
DJIA = x1*Merck + x2*J&J + x3*Pfizer + x4*DuPont + ....
5
Is this chart familiar?
![Page 6: From Design of Experiments to closed loop control · From Design of Experiments to closed loop control Petter Mörée & Erik Johansson Umetrics](https://reader030.vdocuments.us/reader030/viewer/2022041204/5d52c42288c993073e8b712d/html5/thumbnails/6.jpg)
MSPC – Multivariate Statistical Process Control Evolution Level – Monitoring
• Example of a fermentation
-3
-2
-1
0
1
2
3
4
0 10 20 30 40 50 60 70 80 90 100 110 120 130
tPS
[1]
$Time (normalized)
PO_WST3433_EXJADE_Drying_V01.M3:3Predicted Scores [comp. 1]
+3 Std.Devt[1] (Avg)-3 Std.DevtPS[1] (Batch S0058_A_854826)
SIMCA-P+ 11 - 01.08.2009 14:42:24
Control limits
Average (signature) of all good experiments
New run/experiment assessed by the model
t1= x1*Temperature + x2*Pressure + x3*Agitation speed + x4* pO2.
![Page 7: From Design of Experiments to closed loop control · From Design of Experiments to closed loop control Petter Mörée & Erik Johansson Umetrics](https://reader030.vdocuments.us/reader030/viewer/2022041204/5d52c42288c993073e8b712d/html5/thumbnails/7.jpg)
Statistical Process Control MULTIVARIATE CONTROL CHART
-9
-8
-7
-6
-5
-4
-3
-2
-1
0
1
2
-100 0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300 2400 2500 2600 2700
t[1]
$Time (smoothed)
SIMCA-P+ 11 - 14.03.2011 17:53:19
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
8400 8500 8600 8700 8800 8900 9000 9100 9200 9300 9400 9500 9600 9700 9800 9900 10000 10100
Num
0 008 6856 5 30 ceto e_ 3 0e 3
Multivariate Process
Signature
average of all good runs control limits (± 3σ from avg.)
-20
-10
0
10
20
-80 -70 -60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60 70 80
t[2]
t[1]
Model Data Ciclo - Oct 2010 v5 - batch level (scores).M2 (PCA-X)t[Comp. 1]/t[Comp. 2]
R2X[1] = 0.873728 R2X[2] = 0.0564179 Ellipse: Hotelling T2 (0.95)
612
614 615
616
617
618
619
620
621
622
623
624625
626 627
628
633
634
630
631
632
636637
638
635
639
641
SIMCA-P+ 11 - 10.03.2011 19:27:42
![Page 8: From Design of Experiments to closed loop control · From Design of Experiments to closed loop control Petter Mörée & Erik Johansson Umetrics](https://reader030.vdocuments.us/reader030/viewer/2022041204/5d52c42288c993073e8b712d/html5/thumbnails/8.jpg)
MVDA Objectives for the pharmaceutical & biopharmaceutical industry
• Increase of process understanding – Identification of influential process parameters – Identification of correlation pattern among the process parameters – Generation of process signatures – Relationship between process parameters and quality attributes
• Increase of process control – Efficient on-line tool for
• Multivariate statistical control (MSPC) • Analysis of process variability
– Enabling on-line early fault detection – Support for time resolved design space verification
• real time quality assurance – Predicting quality attributes based on process data – Excellent tool for root cause, trending analysis and visualization – Fundament for Continued Process Verification (CPV)
Developm
ent Production
![Page 9: From Design of Experiments to closed loop control · From Design of Experiments to closed loop control Petter Mörée & Erik Johansson Umetrics](https://reader030.vdocuments.us/reader030/viewer/2022041204/5d52c42288c993073e8b712d/html5/thumbnails/9.jpg)
Work and Data flow For Method Development
All Process Parameters
Evolution Level
Batch Level
Individual Probes
Individual Probes …
Reduction of Dimensionality
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
320 330 340 350 360 370 380 390 400 410 420
Num
Recorded Process Parameter during granulationObsID(Obs ID ($PhaseID))Mixer Power rate of change precss variable0.01 * Mixer torqute process variable0.1 * Mixer speed process variable0.1 * Product temperature process variableMixer power process variavle (electrical)
Aims: - Creation of batch signature - Identify correlation patterns - Fundament for CPV
![Page 10: From Design of Experiments to closed loop control · From Design of Experiments to closed loop control Petter Mörée & Erik Johansson Umetrics](https://reader030.vdocuments.us/reader030/viewer/2022041204/5d52c42288c993073e8b712d/html5/thumbnails/10.jpg)
Work and Data flow For Routine Use in Production
Identification of responsible
Parameter(s)
Evolution Level
Batch Level
Investigation on process data
Aims: -Conformity check - Real time release testing - Trend analysis - Root cause analysis
-2
-1
0
1
2
Flo
w liq
uid
feed
Mix
er
Pow
er
Rate
Mix
er
pow
er
Mix
Pow
er
(calc
ula
ted)
Mix
er
Torq
ue
Liq
uid
feed p
um
p s
peed
Bow
l P
ressure
Mix
er
Speed
Chopper
Speed
Pro
duct T
em
pera
ture
Bow
l T
em
pera
ture
Tota
l Liq
uid
Added
Score
Contr
ibP
S(S
0007_B
_854825:1
5.7
895 -
Avg:1
5.7
895), W
eig
ht=
p1
Var ID (Primary)
PO_WST10332_EXJADE_GRAN_Steintraining.M2:7, PS-ComplementoryScore Contrib PS(S0007_B_854825:15.7895 - Avg:15.7895), Weight=p[1]
Mis
sin
g
SIMCA-P+ 11 - 08.02.2009 17:17:00
-40
-20
0
20
40
60
80
100
120
0 10 20 30 40 50
$Time (normalized)
PO_WST10332_EXJADE_GRAN_Steintraining.M2:7Predicted Liquid feed pump speed
+3 Std.DevXVar(Liquid feed pump speed) (Aligned) (Avg)-3 Std.DevXVarPS(Liquid feed pump speed) (Batch S0007_B_854825)
SIMCA-P+ 11 - 08.02.2009 17:17:46
-180
-160
-140
-120
-100
-80
-60
-40
-20
0
20
40
-20 -10 0 10 20 30 40 50 60 70 80 90 100 110 120
tPS
[2]
tPS[1]
PO_WST10332_EXJADE_GRAN_Steintraining - batch level (scores).M1 (PCA-X), All Batches, PS-ComplementorytPS[Comp. 1]/tPS[Comp. 2]
R2X[1] = 0.402627 R2X[2] = 0.341738 Ellipse: Hotelling T2PS (0.95)
S0006_A_85S0007_A_85S0007_B_85
S0006_A_85S0006_B_85
S0007_A_85S0007_B_85S0008_A_85
S0008-B_85S0009_A_85S0009_B_85
S0010_A_85S0010_B_85S0011_A_85S0011_B_85
S0012_A_85S0012_B_85S0014_A_85S0014_B_85S0018_A_85S0018_B_85S0019_A_85S0019_B_85S0021_A_85S0021-B_85
S0022_A_85S0022_B_85S0023-A_85S0023-B_85S0024-A_85S0006_B_85S0025_A_85S0025_B_85S0026_A_85S0026_B_85S0027_A_85S0027_B_85S0028_A_85
S0028-B_85
S0029-A_85S0029-B_85S0030-A_85S0030_TEILS0031_A_85S0031_B_85S0032_A_85S0032_B_85S0033_A_85S0033-B_85S0034_A_85S0034_B_85S0035_A_85S0035_B_85S0037_A_85S0037_B_85S0039_A_85S0039_B_85
S0040_A_85S0040_B_85S0041_A_85S0041_B_85S0042_A_85S0042_B_85S0043_A_85
S0044_A_85S0044_B_85S0045_A_85S0045_B_85S0046_A_85S0046_B_85S0047_A_85S0047_B_85S0048_A_85S0048_B_85
SIMCA-P+ 11 - 08.02.2009 17:15:01
Increased of level of detail Answers: What? When? How?
![Page 11: From Design of Experiments to closed loop control · From Design of Experiments to closed loop control Petter Mörée & Erik Johansson Umetrics](https://reader030.vdocuments.us/reader030/viewer/2022041204/5d52c42288c993073e8b712d/html5/thumbnails/11.jpg)
What makes Multivariate-SPC so powerful?
• The SIMCA product family uses a data compression technique
– Multivariate data analysis • PCA and or PLS
• Data from all relevant process parameters are concentrated to a few highly informative graphs
– Simplifies overview, analysis and interpretation
– Enable use of data by increasing ease of use
• Simple drill-down functionality to transfer compressed information back to raw data for analysis
![Page 12: From Design of Experiments to closed loop control · From Design of Experiments to closed loop control Petter Mörée & Erik Johansson Umetrics](https://reader030.vdocuments.us/reader030/viewer/2022041204/5d52c42288c993073e8b712d/html5/thumbnails/12.jpg)
Drill-down for analysis
![Page 13: From Design of Experiments to closed loop control · From Design of Experiments to closed loop control Petter Mörée & Erik Johansson Umetrics](https://reader030.vdocuments.us/reader030/viewer/2022041204/5d52c42288c993073e8b712d/html5/thumbnails/13.jpg)
Monitor
• Early fault detection – SIMCA-online technology is
acknowledged for its ability to detect process issues before they become critical
• Project dashboard – Full drill-down to raw data for
cause analysis
• Knowledge building – Instant analysis of process
changes improves understanding
• Process visibility – Easy-to-grasp graphics makes
the process status accessible to colleagues at all levels
![Page 14: From Design of Experiments to closed loop control · From Design of Experiments to closed loop control Petter Mörée & Erik Johansson Umetrics](https://reader030.vdocuments.us/reader030/viewer/2022041204/5d52c42288c993073e8b712d/html5/thumbnails/14.jpg)
Prediction and Continued Process Verification
• Product quality information – Indirect information based on
process behavior – As long as a process behaves
well, product should be according to specification
• Soft sensor modeling – Predict hard-to-get process
properties from online process data, spectral data etc.
• Predictive analytics – Online prediction of product
quality and properties
• Continued Process Verification – Ongoing assurance is gained
during routine production that the process remains in a state of control.
![Page 15: From Design of Experiments to closed loop control · From Design of Experiments to closed loop control Petter Mörée & Erik Johansson Umetrics](https://reader030.vdocuments.us/reader030/viewer/2022041204/5d52c42288c993073e8b712d/html5/thumbnails/15.jpg)
Motivation for QbD
• Reducing process variability is not necessarily desirable
00.20.40.60.8
1
Input
00.20.40.60.8
1
Process
00.20.40.60.8
1
Output
With variation in inputs • Initial material qualities • Environment • Equipment
Static process Results in variability in outputs
15
![Page 16: From Design of Experiments to closed loop control · From Design of Experiments to closed loop control Petter Mörée & Erik Johansson Umetrics](https://reader030.vdocuments.us/reader030/viewer/2022041204/5d52c42288c993073e8b712d/html5/thumbnails/16.jpg)
QbD and PAT Strategies
• Control strategy b) feedforward control
00.20.40.60.8
1
Input
00.20.40.60.8
1
Process
00.20.40.60.8
1
Output
Adjusting the process based on variations in the input • Media and feed composition • Used in pulp and paper and other industries with natural products
with high variability • Cheese production
16
![Page 17: From Design of Experiments to closed loop control · From Design of Experiments to closed loop control Petter Mörée & Erik Johansson Umetrics](https://reader030.vdocuments.us/reader030/viewer/2022041204/5d52c42288c993073e8b712d/html5/thumbnails/17.jpg)
QbD and PAT Strategies
• Control strategy c) PAT control
00.20.40.60.8
1
Input
00.20.40.60.8
1
Process
Adjusting the process based on measurement of quality in the process • Used in many processing industries using various methods
• Direct measurement of material quality • Inferential control – estimation of quality from process
measurements • Spectral calibration
00.20.40.60.8
1
Output
17
![Page 18: From Design of Experiments to closed loop control · From Design of Experiments to closed loop control Petter Mörée & Erik Johansson Umetrics](https://reader030.vdocuments.us/reader030/viewer/2022041204/5d52c42288c993073e8b712d/html5/thumbnails/18.jpg)
18 UMETRICS CONFIDENTIAL
• Monitoring is used to detect and diagnose process deviations
Monitoring
Important Process Parameter
![Page 19: From Design of Experiments to closed loop control · From Design of Experiments to closed loop control Petter Mörée & Erik Johansson Umetrics](https://reader030.vdocuments.us/reader030/viewer/2022041204/5d52c42288c993073e8b712d/html5/thumbnails/19.jpg)
19 UMETRICS CONFIDENTIAL
• MPC is used to predict
Model Predictive Control (MPC)
Important Process Parameter
![Page 20: From Design of Experiments to closed loop control · From Design of Experiments to closed loop control Petter Mörée & Erik Johansson Umetrics](https://reader030.vdocuments.us/reader030/viewer/2022041204/5d52c42288c993073e8b712d/html5/thumbnails/20.jpg)
20 UMETRICS CONFIDENTIAL
• MPC is used to predict and optimize the process
Model Predictive Control (MPC)
Important Process Parameter
![Page 21: From Design of Experiments to closed loop control · From Design of Experiments to closed loop control Petter Mörée & Erik Johansson Umetrics](https://reader030.vdocuments.us/reader030/viewer/2022041204/5d52c42288c993073e8b712d/html5/thumbnails/21.jpg)
21 UMETRICS CONFIDENTIAL
Model Based Control
Manipulated Variables
Important Process Parameter
![Page 22: From Design of Experiments to closed loop control · From Design of Experiments to closed loop control Petter Mörée & Erik Johansson Umetrics](https://reader030.vdocuments.us/reader030/viewer/2022041204/5d52c42288c993073e8b712d/html5/thumbnails/22.jpg)
Novartis Biopharmaceutical
22
![Page 23: From Design of Experiments to closed loop control · From Design of Experiments to closed loop control Petter Mörée & Erik Johansson Umetrics](https://reader030.vdocuments.us/reader030/viewer/2022041204/5d52c42288c993073e8b712d/html5/thumbnails/23.jpg)
23
![Page 24: From Design of Experiments to closed loop control · From Design of Experiments to closed loop control Petter Mörée & Erik Johansson Umetrics](https://reader030.vdocuments.us/reader030/viewer/2022041204/5d52c42288c993073e8b712d/html5/thumbnails/24.jpg)
![Page 25: From Design of Experiments to closed loop control · From Design of Experiments to closed loop control Petter Mörée & Erik Johansson Umetrics](https://reader030.vdocuments.us/reader030/viewer/2022041204/5d52c42288c993073e8b712d/html5/thumbnails/25.jpg)
![Page 26: From Design of Experiments to closed loop control · From Design of Experiments to closed loop control Petter Mörée & Erik Johansson Umetrics](https://reader030.vdocuments.us/reader030/viewer/2022041204/5d52c42288c993073e8b712d/html5/thumbnails/26.jpg)
![Page 27: From Design of Experiments to closed loop control · From Design of Experiments to closed loop control Petter Mörée & Erik Johansson Umetrics](https://reader030.vdocuments.us/reader030/viewer/2022041204/5d52c42288c993073e8b712d/html5/thumbnails/27.jpg)
27
Chemometric portfolio
![Page 28: From Design of Experiments to closed loop control · From Design of Experiments to closed loop control Petter Mörée & Erik Johansson Umetrics](https://reader030.vdocuments.us/reader030/viewer/2022041204/5d52c42288c993073e8b712d/html5/thumbnails/28.jpg)
Thank you for your attention!
28