Product Release: Unscrambler® X Process Pulse II

0

Unscrambler® X Process Pulse IIWe are pleased to announce the release of a new version of our easy-to-use multivariate process monitoring solution, the Unscrambler® X Process Pulse II.

The new version provides users with many new and improved features, including parallel import and alignments of multiple data sources in real time, running and visualization of models with real time flagging of deviations, as well as data historian for simplified troubleshooting. The new version has also been updated with a new and more intuitive user interface.

Unscrambler® X Process Pulse II is used in a wide range of industries and research fields, for improvement in product development, manufacturing and quality control. Using powerful multivariate models, the software helps manufacturers and process operators identify and correct deviations before they become problems. The product includes predictions, classifications and projections, presented in an intuitive graphical interface, and is easily integrated with third-party production and processing systems.

Key business benefits that can be achieved from using Unscrambler® X Process Pulse II include:

  • Reduce process failures with Early Event Detection
  • Improve yields through better process understanding
  • Reduce raw material, scrap, energy and rework costs
  • Speed up cycle time and run processes closer to limits
  • Accelerate scale up from R&D to production scale

To learn more about Unscrambler® X Process Pulse II and watch in action, please take a look at the webinar recording here.

In the coming months we will also arrange training courses for Unscrambler® X Process Pulse II, please watch our training section for updates.

Take a look at our new Unscrambler® X Process Pulse II introduction video!

Tech Tip: Combining Data Matrices

0

Often, during the initial setup of a project (or when revisiting one), there is a need to combine data matrices. It may be that new data has been collected that needs to be added to a calibration set. Or perhaps you are combining data sources for the same samples (process data, spectroscopic data, analytical tests, etc). In simple cases, you can always copy/paste the desired information, but if you have more than 2-3 matrices to combine, this can be a hassle. A nice, quick alternative is to use the Matrix Calculator function in The Unscrambler® X. This example will show combining sample sets.

1) Import all the data that you are interested in combining, and label your data matrices appropriately.
2) Check that number of variables/samples is consistent with each matrix. If appending new samples, then the number of variables must be the same and in the same order. If augmenting new data for existing samples, then the number of samples must be the same and in the same order. Note stay tuned for our upcoming sample alignment tool!
3) Select Tools – Matrix Calculator
4) Check the boxes next to each of the matrices you want to combine. And click on the Shaping tab.
Optionally, you may select Add category variable, which will add a new column with a category variable labeling which matrix the data came from. This is especially helpful when combining batches or different raw material types (such as in this example).
5) Click on Append to combine the data sample-wise. Note: clicking on Augment would combine the data variable-wise. Notice it is not currently available because the matrices do not have the same number of samples.
6) Click Close, and check your new combined matrix. If you clicked Add category variable, you will see that appended to the end of the matrix.

Tech Tip: Group Rows

0

Sometimes when you have data that can be split into different categories (such as classes or batches), it is helpful to create row sets based on those categories. Having row sets identified can make it easier to search for grouping in scores plots, identifying outliers (especially mislabeling), creating classification models, or creating individual models when a global model is inappropriate. There is a function in Unscrambler that is very handy for doing this automatically if you have a Category variable set up.

In this example, there are a number of samples of different types of vegetable oils. There is a category variable that identifies each sample as one of the available types.

data1

To define row sets automatically based on this category variable, select/highlight the column of interest, then navigate to Edit àGroup Rows…

data2

A dialogue box will appear and you can confirm or edit your column selection.

Note: If the column has numerical data in it, the “Number of Groups” option will be functional, and the software will automatically create evenly distributed ranges for the number of groups selected.

data3

Once “OK” is pressed, the appropriate row sets (either based on category values or automatically generated ranges) will be defined and available in the navigation pane on the left side of the Unscrambler window.

data4

CAMO Software announces a strategic partnership with Stat-Ease, Inc.

0

Dear friends and colleagues,

We are very pleased to announce that CAMO Software has entered into a strategic partner agreement with Stat-Ease, Inc., a leading provider of design of experiments (DoE) software and services. CAMO will bundle Stat-Ease’s Design-Expert® software with CAMO’s The Unscrambler® X. Customers will have access to two world-leading software packages providing advanced data analysis in one single integrated solution.

At CAMO, we are very excited about this collaboration. It will provide CAMO with a broader product and service offering to new and existing clients, and offer Stat-Ease a global extension to their sales and marketing organization.

We are also very pleased to announce that Pat Whitcomb, founder and president of Stat-Ease, Inc., and Mark Anderson, principal of Stat-Ease, Inc., will be joining us at our User Meeting in Prague in April.  They have kindly agreed to give a presentation about managing uncertainty in a design space and how to perform effective experimentation.

Click here to read the updated agenda and to sign up for the User Meeting.

To learn more about the new solution, please contact your CAMO Account Manager.

To learn more about Stat-Ease and Design-Expert® please click here.

CAMO User Meeting 2015, Prague

0

Join us at this exciting event to learn from scientific and industry leaders how you can improve processes and products with the power of multivariate analysis and process monitoring software.

This informative event is a great opportunity to hear from world-leading experts, share experiences with other professionals and have one-on-one meetings with the CAMO team.

We are also pleased to announce some of our esteemed speakers, please see below. More will be announced shortly.

WHO SHOULD ATTEND
The User Meeting is tailored for people working in:

  • Product development/R&D
  • Production/Manufacturing
  • Technical Services or PAT/QbD groups
  • Quality Assurance and Control

The event is relevant for users of The Unscrambler® software suite as well as people new to multivariate analysis and the process monitoring arena who are interested in learning how it can improve product development, manufacturing processes and quality control.

AGENDA OVERVIEW
More details will be provided in the coming weeks

  • Keynote talks from world-leading experts within chemometrics/multivariate data analysis from both science and industry including Professor Harald Martens from NTNU, Dr. Waltraud Kessler from Hochshule Reutlingen, Dr. Tobias Merz from Lonza AG and Manfred Dausch from Dausch Technologies
  • Exclusive demonstrations of major new products to be released by CAMO Software
  • Presentations from The Unscrambler® X and Unscrambler® Process Pulse customers with applications covering various industries
  • Panel discussion with CAMO experts and industry leaders
  • In-depth sessions with CAMO’s leading scientific and product experts

COSTS
Option 1: Full conference package: €460
Includes 1 night accommodation, breakfast, lunch both days and dinner with city tour on Thursday night.

Option 2: Conference only package: €335
Includes breakfast, lunch both days and dinner with city tour on Thursday night.
People intending to stay overnight from Wednesday 15th or over the weekend should contact CAMO directly to arrange the additional accommodation.

SPRING IN PRAGUE
Prague is one of Europe’s most fascinating and historic cities, with a history dating back over 1100 years. It offers something for everyone, with famous sights such as the Vysehrad Castle, The Charles Bridge and the Museum of Decorative Arts, great restaurants and bars, and a combination of both old and modern architecture. The city comes alive in spring, making it the ideal time to enjoy Prague’s wonderful riverside walks.
Read more about Prague

DATES & TIMES
Thursday 16th April
10am – 5pm
(dinner and tour following)

Friday 17th April
9am – 3pm

Find out more here…

Release of Unscrambler© X Mie Plugin v1.0

0

Mie scattering can be observed in infrared spectra of single cells measured by high energy (Synchrotron) spectroscopy.
We are pleased to announce the release of a new plugin that can be used to correct for the Mie scattering in The Unscrambler© X. The plugin generates Mie contribution spectra that can be used as input to the Extended Multiplicative Signal Correction (EMSC) transform in order to remove the Mie effects from the raw spectra.

The details of the used algorithm have also been published in:
A. Kohler et al. Estimating and Correcting Mie Scattering in Synchrotron-Based Microscopic
Fourier Transform Infrared Spectra by Extended Multiplicative Signal Correction, Appl.
Spectrosc., 62, 259-266 (2008).

The plugin is available to all users with a valid subscription to the CAMO Support and Maintenance Program, and is available for download here. For more information and installation instructions, please refer to the release notes.

PAT success

0

PAT (Process Analytical Technology) is a tool within the QbD (Quality by Design) toolbox enabling real-time quality assurance. However the use of PAT is still not standard within the life sciences. The reasons may be many but typical explanations are that it is too time consuming, too costly, or there is simply no proof that it will benefit the operations. The premise for any PAT implementation is that it has to benefit the operations in an efficient and economically meaningful way.

The two fundamental requirements for successful PAT projects are relevant and cost effective tools in, addition to a justified business case.  CAMO Software is working in this space to offer customers market leading software solutions for PAT applications.

The Unscrambler X is regarded as the best tool for analysis of historical PAT data with its wide selection of Chemometrics methods and intuitive user interface. CAMO also offers Process Pulse, which makes using Unscrambler models in online PAT applications easy. Process Pulse is scalable from the single user in a local laboratory to the fully connected enterprise version for the entire organization.

Recently CAMO went into an agreement with Lonza to provide them with an adapted enterprise version of Process Pulse. The solution connects directly to multiple data sources in the Lonza production lines and records and stores the data in secure databases. During data recording multivariate models can be run to track process performance and quality parameters, enabling real-time quality control. The recorded data is also available for future data analysis, troubleshooting and product traceability.

Data gathering and evaluation is becoming more important in the total life-cycle of a product and not just in the commercial scale production. Building processes start in the R&D labs and should be validated in the production plants. To close the gap between lab scale and production scale, the PAT data management software should be able to bring all data into one common platform.

The time consuming manual work will be supported and fully automated by the Process Pulse software. The main aspects are:

  •  Merging of multivariate and univariate data sets
  • Bring data to a common time axis
  • Identification of Batches
  • Cleaning up of data sets
  • Detection of interessting process steps
  • Creation of Overlays

One example of combined data from lab scale and production scale shows the figure 1. Due to the overlay of all batch trajectories the variance and the scaling effect can easily be demonstrated.


Figure 1: Combined Information of small scale and large scale batch trajectories in the PCA Score coordinate system.

The integration of a PAT data management system into the development workflow is an important step to close the gap between R&D and production. And, process understanding becomes more integrated in daily business.

This Application Note was prepared by Dr. Geir Rune Flåten and co-authored by Dr. Tobias Merz, PAT Team Leader at Lonza AG.

MULTIVARIATE DATA ANALYSIS WITH THE UNSCRAMBLER®

0

Price:    Conferee $450; Student $350; Nonconferee $450
Date:    Monday, September 29, 9:00 am 4:30 pm
Instructor:    Heather Brooke, CAMO Software

This course is ideal for people who are:

  • new to multivariate analysis, those wanting a refresher on methodology
  • process engineers looking at strategies for implementing multivariate statistical process control
  • working with sensory or other survey based data
  • spectroscopists who want to have a better understanding of the fundamentals of multivariate data as applied to other non-spectroscopic situations

Course outline

  • What is Multivariate Analysis
  • Motivation for using Multivariate Analysis
  • Data Collection and Visualization
  • Exploratory Data Analysis
  • Outlier detection using PCA
  • An introduction to Multivariate Regression
  • Interpretation of Multivariate Regression Models
  • Correct Validation Practices
  • Summary and Overview

Read more http://www.camo.com/news-events/index.html

 

Einladung zum kostenlosen Seminar: Versuchsplanung und Multivariate Datenanalyse für die Produktentwicklung und Echtzeit Qualitätskontrolle

0

CAMO Software AS als einer der führenden Anbieter von Software-Lösungen für die Multivariate Datenanalyse wird in Deutschland eine Reihe von Seminaren abhalten, um Ihnen nützliche Informationen zur Versuchsplanung und zur Multivariaten Datenanalyse zu geben. Diese Seminare werden in München, in Frankfurt, Düsseldorf und Hamburg durchgeführt. Sie werden erfahren, wie Sie Ihre Daten mit optimalem Nutzen analysieren.

Unsere Experten werden Ihnen anhand von “Best-Practice Methoden” aufzeigen, wie Sie noch gezielter auf steigende Anforderungen in der Multivariaten Welt reagieren können.

Die Teilnahme ist kostenlos, eine leichte Mahlzeit im Anschluss und Getränke werden serviert.

Datum: Dienstag, 14. Oktober 2014
Uhrzeit: 10:00 – 13:30 Uhr
Ort: München – Ramada Hotel & Conference Center, Konrad Zuse Platz 14, 81829 München
Registrieren Sie sich hier

Datum: Mittwoch, 15. Oktober 2014
Uhrzeit: 10:00 – 13:30 Uhr
Ort: Frankfurt – Lindner Hotel & Sports Academy, Otto-Fleck-Schneise 8, 60528 Frankfurt
Registrieren Sie sich hier

Datum: Donnerstag, 16. Oktober 2014
Uhrzeit: 10:00 – 13:30 Uhr
Ort: Düsseldorf – Hyatt Regency, Speditionsstrasse 19, 40221 Düsseldorf
Registrieren Sie sich hier

Datum: Freitag, 17. Oktober 2014
Uhrzeit: 10:00 – 13:30 Uhr
Ort: Hamburg – Sofitel Hamburg Alter Wall 40, 20457 Hamburg
Registrieren Sie sich hier

Für die Registrierung und für mehr Informationen, kontaktieren Sie bitte unseren Country Manager
Jens J. Oestreich:
Email: jo@camo.com
Tel: +47 22 39 63 00

Wir freuen uns auf Ihre Teilnahme!
Mit freundlichen Grüssen
CAMO Software AS
Oslo, Norwegen

 

Some words about non-linearity in the context of multivariate methods

0

Dr. Frank Westad, CAMO Software

The latent variable methods PCA, PCR and PLSR are linear, or more correctly bilinear as they are linear in both scores and loadings. Many real-world processes are inherently non-linear in one way or another. The non-linearity can be in terms of the overall relationship between two sets of variables (X, Y) or on a more individual basis for some specific variables based on first-principle or deterministic models. Both curvature and more or less known underlying phenomena in the system might introduce non-linearity in the data. The non-linear effects can be handled in different ways:

1.    Pre-processing with the purpose of removing non-linearity
2.    Transform variables based on a priori knowledge
3.    Include interaction and square terms(and optionally higher order terms) as a rough approximation
4.    Use non-linear methods

Non-linear methods have been developed for most linear and bilinear regression methods. In the case of PLS regression two sets of scores are computed for every a factor; the x-scores, ta, and the y-scores, ua, which are the basis for the so-called inner relation which is found by the least squares solution. An alternative is to perform non-linear inner relation modelling (polynomial functions, spline functions). Artificial neural networks (ANN,) have been investigated thoroughly over the past years as a non-linear family of methods. More recently Support Vector Machines for classification and regression have shown good figures of merit. Other methods aim at making local linear models, e.g. Locally Weighted Regression.

Below is a suggested general procedure for how to include terms that express non-linearity within the family of linear and bilinear methods:

1.    Make an initial model with the linear terms and relevant variable transformations due to a priori knowledge, e.g. Viscosity = f(Temperature2).
2.    Decide on significant variables from some suited test of significance, e.g. cross-validation at the correct scientific level (see below).
3.    Recalculate with significant variables and include interaction terms and square terms (I & S). Remove non-significant variables.
4.    Recalculate with the significant variables only.

The term “significant” might be replaced by the more loose term “relevant” when aspects other than statistical significance also are part of the data analysis.

Validation is even more essential when performing non-linear modelling, as random non-linear tendencies and outlying objects might be modelled as “true” non-linear relations. The objects in a data table can often be stratified into groups based on background information about the origin of the objects. Such groups are a consequence of the experimental set-up of the study. Typical stratifications are:

–    Across instrumental replicates (repeatability)
–    Reproducibility (analyst, instrument, reagent…)
–    Sampling site and time
–    Across treatment/origin (year, raw material batch, lot ID…)

Cross-validation (CV) performed at the various grouping level will give important information about the stability of the model and which sources of variation that need special attention. Thus, even if a test set has been defined as the proper way of validating the model (or process or system in a wider context) the calibration set must be validated with CV at the appropriate level. If not, the model dimensionality may not be conservative enough and the test set is predicted with a suboptimal number of variables or factors.

Some additional comments regarding item 3 above: It is rare that a variable is not significant as a linear term when the “true” relation is somewhat non-linear, except from a pure “bell-shaped” relation. This will be revealed in interpretation from the available model plots such as correlation loadings, so the danger of losing vital information is rather low. The score plot may change its appearance when I & S terms are included, and the model can have lower explained validation variance than the first linear model. The model performance will usually improve when only the significant variables are used in the recalculation. Validation becomes even more important for non-linear methods and when non-linear terms for individual variables are added, as the danger of overfitting increases and outliers in a linear model may be modelled as a non-linear trend.

Although PCR and PLSR are linear methods, they have shown to be suitable also for data with inherent non-linearity. This is sometimes at the cost of using 1-2 factors more than a direct non-linear method.

Back to Top Copyright © 2015 CAMO Software. All rights reserved.