Short Summary
The course is aimed at individuals who work with batch processes, who want to understand the sources of variation in batch processes and the impact that these variations may have on product quality, and who want to use the latent variable models for setting up batch monitoring schemes, for predicting and controlling product quality, and for optimizing and scaling-up batch processes.
All of the methods are illustrated through applications to industrial processes. Attendees will have the opportunity to analyze many of these industrial data sets in several workshop sessions using multivariate software specifically developed for these problems.
Background
Large amounts of data are routinely collected around batch processes; however, better use of these data can be made for design, scale-up, operation, monitoring, and control of batch processes throughout their life cycle.
This course describes methods which can be used to achieve the following:
- Generate process knowledge and understanding by identifying which variables and sources of variation in the process have important influences on final quality attributes.
- Integrate data from on-line instrumentation, from traditional on-line process measurements and from discrete samples into the analysis, monitoring and control of batch processes.
- Integrate knowledge from processes at multiple scales (bench scale, pilot plant, full scale) for use in process design, qualification and verification.
- Establish on-line statistical monitoring schemes to detect the presence and degree of any variation in the process and to validate the continuing performance of the process.
- Design process control schemes that will mitigate (reduce or compensate for) the impact on quality of both external (raw material and environmental) variations and internal batch variations.
The latent variable methods, PCA (principal component analysis) and PLS (projection to latent structures) are used to achieve these goals.
The course is structured with an equal blend of theory, example applications and computer labs.
Target Audience
This intensive short course will be relevant to scientists, engineers and statisticians who regularly deal with process data. It is highly relevant to the chemical, food, semi-conductor, biotech and pharmaceutical industries. In particular, given the new proposed FDA guidelines for the validation of batch pharmaceutical processes, it is particularly timely for people dealing with the development and validation of batch processes in this industry.
Prerequisites
This course is aimed at engineers, scientists and statisticians who work with batch processes and have had an initial introduction to the use of multivariate methods based on PCA and PLS.
Materials We Provide
All the material will be available in a full-colour binder for participants to take notes in. You will also receive a CD that contains the course notes in electronic format (print-quality PDF), the example data sets and additional papers with much more details on the methods. A fully functional version of our software package, ProSensus MultiVariate, is also provided.
Software and Computer Sessions
This course uses interactive computer sessions, using a number of industrial data sets, to reinforce the concepts. ProSensus focuses on teaching the concepts and practical use of the methods, with real data sets.
The software used in the course is our package, Prosensus MultiVariate. This package has all the features of the major multivariate packages on the market, and more! However, participants are welcome to use an alternative software package if preferred. Participants are encouraged to bring their laptops with any other multivariate software packages installed.
Also, we strongly recommend that participants bring their own industrial data for use during the computer workshops. [Text format or spreadsheet files, eg: Excel, are preferred.]
Course Content
Please see this page to view the detailed course content.
Your Instructors
John MacGregor is the President of ProSensus and a professor in chemical engineering at McMaster University. He holds M.Sc. degrees in both chemical engineering and statistics, and a Ph.D. degree in statistics from the University of Wisconsin. Dr. MacGregor has worked as an engineering specialist at Monsanto, and a consultant in the areas of advanced process control, statistical process control, and polymer reaction engineering, to many companies in North America, Europe and beyond.
He has received many awards: he has been inducted as a Fellow of the Royal Society of Canada and a Fellow of the Canadian Academy of Engineering for his work in multivariate methods for process monitoring, analysis and optimization. The Canadian Society for Chemical Engineering (CSChE) awarded him a Century of Achievement Award, for his significant industrial and academic accomplishments. He was also the first person outside Scandinavia to be awarded the Herman Wold Gold Medal by the Swedish Chemometrics Society for his work in multivariate projections and analysis. He is a fellow of the American Statistical Association (ASA), and recipient of the W.G. Hunter Award and the Shewhart Medal from the American Society for Quality (ASQ).
Other portions of the course will be given by ProSensus employees, who will also assist with the computer laboratories.
Course Cost and Materials
The cost per participant is US$1990 per person, which includes the price of all materials. A discount of $100 per person is given to companies that register two or more participants, i.e. two or more participants means $1890 per person.
Academic discounts are also available. Please contact ProSensus for further details.
The materials include all presentation slides (printed in colour and the electronic PDF version), all data sets used, and additional papers with more details on the multivariate methods.
Location and Accommodation
The course will be held in Burlington, Ontario at the Waterfront Hotel. Special room rates have been arranged with the resort (standard room $125+HST). To get these rates you must book 30 days in advance through their 866-685-3457 number and state that you are with the ProSensus course. Details and photos of the hotel are available here.
