APACT '19 Programme

Programme

Pre-conference courses

Four optional half day pre-conference courses will run in parallel on Monday 29th April 2019 at the Crowne Plaza Hotel.

These courses will be open to both conference delegates as a low cost additional option, or may be attended on a ‘course only’ basis.


COURSE 1: DRIVING YOUR PROCESS MONITORING PROGRAMME WITH DATA ANALYTICS

Across industries one of the main challenges in manufacturing processes is to monitor and control variation. Successful variation control ensures predictable production and accompanying business benefi ts. Recently there have been major developments in sensor technology allowing and increasing the number of inline and online measurements. Together with improvements in IT infrastructure and technology, the sensor development contributes to rapid growth in the data available for any process. The big data challenge requires suitable timely analytics to enable each factory to make the most of the opportunities that the new paradigm is offering. A common label for this challenge is Industry 4.0.

Analytics is a key component within Industry 4.0. Relevant actionable results are required, and these results must be available for the control system in a timely manner. Multivariate methods are well established and proven useful within process monitoring. Lately there have also been a number of applications where machine learing and artifi cial intelligence (AI) have been considered. The choice of method depends on the type of data available and the purpose of the analytics.

In this training, the challenges and opportunities for monitoring and control within the big data paradigm are reviewed. The focus is on analytics and integration within the manufacturing process. The underlying question which we are trying to answer is how to create and apply manufacturing intelligence.


11.45

Registration and coffee


12.00

Lunch


13.00

Industry 4.0 and analytics; Analytical strategies - trending and modelling; Which methods for which data


15.00

Tea/coffee


15.20

Multivariate analysis and machine learing and AI; Process monitoring program - what, how and when; Manufacturing intelligence


This course will be facilitated by Geir Rune Flaten, CAMO Analytics.


COURSE 2: SCREENING AND OPTIMISATION DESIGN OPTIONS FOR DOE


11.45

Registration and coffee


12.00

Lunch


13.00

Theory: Introduction to Design of Experiments (DoE). Introduction to designs including: Full & Fractional Factorial (classical), Plackett-Burman, Response Surface, Mixture and Defi nite Screening Designs. The theory behind some of the most commonly applied designs from a design and analysis perspective will be explored with practical examples. Advantages and disadvantages of the designs with a particular focus on Defi nitive Screening Designs will be discussed. Course delegates will share their knowledge and any industrial case studies with the rest of the learning cohort.


15.00

Tea/coffee


15.20

Application: Setting up and analysing designs within software. Course delegates will practise using the designs within a statistical software package. The session will close with a summary discussion and action planning.


This course will be facilitated by Matthew Linsley, University of Newcastle.


COURSE 3: ADVANCED SPECTRAL PREPROCESSING WITHOUT EQUATIONS

Advanced Spectral Preprocessing without Equations takes up where our Chemometrics without Equations course leaves off. It is assumed that participants will have a working knowledge of Principal Components Analysis (PCA) and regression with Partial Least Squares (PLS). The course concentrates on improving chemometric models via advanced preprocessing methods.

The critical difference between inadequate and successful chemometric models is often data preprocessing, i.e. what is done to the data before using PCA, PLS etc. The goal of preprocessing is to remove variation not related to the problem of interest so that the variation of interest is more evident and can be more easily modelled. The course focuses on advanced preprocessing methods, including baselining, normalisation, smoothing and derivatives, Extended Multiplicative Scatter Correction (EMSC) and Generalised Least Squares (GLS), for improving models. The effect of preprocessing on the robustness of the fi nal models is also considered.


11.45

Registration and coffee


12.00

Lunch


13.00

Review of PCA/PLS, centering, normalisation, and baselining methods and specialised scaling


15.00

Tea/coffee


15.20

Filtering and derivatives, GLS, EMSC


This course will be facilitated by Barry M. Wise, President of Eigenvector Research and creator of PLS_Toolbox chemometric software.


COURSE 4: AN INTRODUCTION TO PROCESS CONTROL FOR SCIENTISTS, CHEMISTS AND PHARMACTISTS

The course will provide an overview of closed loop process control for scientists and engineers that have no background in process control and automation. Suffi cient knowledge will be provided to allow delegates to appreciate the scope of process control techniques and the challenges that need to be considered when implementing process control strategies within a PAT and QbD environment. The topics will include classical feedback control with process examples; feedforward and inferential control; multi-loop control; analytical feedback control; and will fi nish with a brief overview of more advanced control systems now being applied. Industrial case studies will be used.


11.45

Registration and coffee


12.00

Lunch


13.00

The Building Blocks of Process Control: An Introduction to Feedback Control, Proportional Control (P); Proportional + Integral Control (PI); Proportional + Integral + Derivative Control (PID) Control


15.00

Tea/coffee


15.20

Multi-loop Control Systems; Cascade Control; Ratio Control; Feedforward Control; Model Based Predictive Control; Inferential Measurement & Software Sensors


This course will be facilitated by Julian Morris, CPACT.

11.00

Registration and exhibition open


11.00

Exhibition


12.00

Lunch break


13.10

Welcome


13.10

Formulation and processing: optimisation manufacturability

Peter Fryer, University of Birmingham, UK

Plenary 1

14.00

Opportunities and challenges for AI in industrial processes and applications

Stephen McArthur, University of Strathclyde, UK

Plenary 2

14.50

New digital technologies and their impact on validation methodology and compliance for GMP and non- GMP applications in life sciences

Lee Hough, Emerson, UK

Keynote 1

15.30

Tea/coffee and Exhibition and Posters


16.00

Application of in-line technology in crystallisation

Andre Powell, AstraZeneca, UK

Keynote 2

16.40

In situ high speed NIR imaging to monitor form change and drug release from rapidly disintegrating tablets

Lucy Hawarden, Bristol-Myers Squibb, UK

Keynote 3

17.20 - 18.20

Exhibitors showcase presentations (including refreshments)


19.00

Dinner


20.00

Social activity

9.00

Chemometrics in PAT: models, measurements and information

Anna de Juan, University of Barcelona, Spain

Plenary 3

9.50

Autonomous calibration and the impact on process analysis and control

Brian Rohrback, Infometrix, USA

Keynote 4

10.30

Tea/coffee and exhibition and posters


11.00

Doing more with less: Semi-supervised calibration with domain-invariant partial least squares regression

Ramin Nikzad-Langerodi, RECENDT, Austria

Parallel session: Multivariate Methods

Data driven process understanding

Gillian Turner, GlaxoSmithKline, UK

Parallel session: Reaction Monitoring

11.30

In-line monitoring of solid sulphate precipitation behaviour in natural and synthetic brines using Raman spectroscopy and MCR

Werner Barnard, Nouryon, The Netherlands

Parallel session: Multivariate Methods

Reaction monitoring of complex systems by NMR spectroscopy – Real examples from the pharmaceutical industry

Steve Coombes, AstraZeneca, UK

Parallel session: Reaction Monitoring

12.00

Combining design of experiments and development of a model predictive controller for industrial scale-up of liquid-liquid formulated product

Min Zhang, University of Birmingham, UK

Parallel session: Multivariate Methods

Performance of inline NIR probes in aqueous polymer dispersions

Jurjen ter Maat, Nouryon, The Netherlands

Parallel session: Reaction Monitoring

12.30

Lunch and Exhibition


13.30

Exhibition and Posters


14.00

Inline spectroscopy: the journey from concept to function

Marion O’Farrell, SINTEF Digital, Norway

Keynote 5

14.40

Online NMR spectroscopy for process analytical applications and aspects on automation and digitization of process industry

Michael Maiwald, BAM, Germany

Keynote 6

15.20

Tea/coffee and Exhibition and Posters


15.50

A comparison of ANNs, SVMs and XGBoost in challenging classifi cation problems

Barry Wise, Eigenvector Research Inc, USA

Parallel session: Machine Learning & Classifi cation Methods

Dynamic drying monitoring of a fi xed bed using 3D-electrical capacitance tomography

Giuseppe Forte, Johnson Matthey Technology Centre, UK

Parallel session: Process Analysis

16.20

Machine learning for pharmaceutical process control and monitoring; Revolution of evolution?

John Mack, Perceptive Engineering, UK

Parallel session: Machine Learning & Classifi cation Methods

Beyond the boundaries of conventional spectroscopy: The benefi ts of high-throughput Raman for process analysis

Mark Kemper, Tornado Spectral Systems, Inc., Canada

Parallel session: Process Analysis

16.50

Detecting valve stiction using classifi ers trained on features extracted from time series

Harrison O’Neill, University of Chester, UK

Parallel session: Machine Learning & Classifi cation Methods

PAT for Industry 4.0: Developments in embedded spectrometer systems and Raman spectroscopy

Christian Lux, tec5 AG, Germany

Parallel session: Process Analysis

17.20

Free time


18.10

Exhibitors showcase presentations (including refreshments)


19.00

Free time


19.30

Conference dinner and entertainment


9.00

The continuous manufacturing control strategy lifecycle

Martin Warman, Martin Warman Consultancy Ltd, UK

Keynote 7

9.40

Integrated control strategy for continuous manufacturing direct compression by using NIR and residence time distribution (RTD) predictive models: Experiences and challenges

Stephen Conway/Juan Rosas, Merck Sharp and Dohme (MSD), UK

Keynote 8

10.30

Tea/Coffee and Exhibition and Posters


11.00

NIR spectroscopy for process characterisation and monitoring of continuous direct compression (CDC)

David Wilsdon, Pfizer, UK


11.30

In-line process monitoring of cell viability and apoptosis in bioreactors using dynamic imaging analysis

Claudia Corredor, Bristol-Myers Squibb, USA


12.00

Next generation manufacturing at Biogen - PAT as an enabler for enhanced process control in commercial antibody manufacture

Oliver Steinhof, Biogen International GmbH, Switzerland


12.30

Poster prize presentation


12.40 - 12.45

Closing remarks, followed by lunch