Process Control aims towards a better production homogenity, a lower number of defective pieces and a better understanding of industrial processes. As methodology to achieve this, we use multivariate data analysis
and modelling.

 

Topics

General Information

Process Control
Process Optimisation
Application Examples
Our offer
Our approach
Our philosophy
Our vision

Multivariate Data Analysis in Process Control
Projects
 
   Process Control

 

Process Control with Multivariate Data Analysis

We are using multivariate data analysis and modelling as basis for process monitoring and control. These methods work in a two-step approach: first, a model is built for the process by implicating the process variables and the output of a process. After thorough validation, the model can be used to visualise and track changes in a process. The difference between actual and optimal values can be fed into the process again for control. The strengths of multivariate data analysis are the genericity of its methods, the almost unlimited number of input variables and its intrinsic ability to handle cross-effects.

Process Optimisation

Another common problem is the optimisation of an industrial process. Often the so-called COST approach (Change One Setting at a Time) is applied for varying the process variables to maximise the yield of the process. However, when the cross-effects of the input variables can not be neglected, the results of this approach will not be optimal. Systematically varying all the variables at the same time and applying multivariate data analysis techniques allows to reach an optimal result with the minimum number of trials. Furthermore, the influence of each variable and their interaction can be estimated.


Application Examples

Actually, multivariate data analysis can be used and is used in almost every field. Following there are application examples of multivariate data analysis and process control in four selected fields.

Food Industry

  • Production process monitoring and control
  • Monitoring of maturity and fermentation processes
  • Determination of ripeness and sweetness of fruits
  • Optimisation of production processes
  • On-line/at-line/in-line measurement of fat, fatty acids, protein, water, sugars, alcohols
  • Determination of geographical origin for anti-fraud action

Petrochemical Industry

  • Production process monitoring and control
  • Quality control
  • Soft sensors

Pulp and Paper Industry

  • Production process monitoring and control
  • Quantification of wood mixtures
  • Measurement of tensile strength
  • Control of Bleach / Measurement of whiteness
  • Monitoring of wastewater and other wastes
  • Measurement of lignin and cellulose
  • Coatings and surface treatment of paper and carton

Economics / Financial Markets

  • Efficiency of strategies: marketing, trading, ….
  • Compound indicators for decision making and control
  • Verification of the effectiveness and ramification of taken measures
  • Evaluation and optimisation of complex business processes

What do we offer?

  • Development of Solutions
    We analyse and evaluate the situation together with your experts. Together we develop a suitable solution. After the roll out we support your team during the full life cycle.
  • Services
    If you already use multivariate techniques, we support you with our expertise. When you are temporary short on resources, we assist your team with our manpower and know-how. Additionally, we can offer deep knowledge in spectroscopy and chemometrics.
  • Training
    To build up knowledge in multivariate data analysis, we offer training for your employees.
  • Consulting
    We offer consulting on all aspects of multivariate data analysis especially for process control, e.g. evaluation of a planned project, selection of the best-suited software for a problem, implementation of modified methods et cetera.
Our approach

We focus on all aspects of multivariate data analysis for process control including the measurements and EDP framework. However, knowledge and experience is needed in both the methodology itself and the process under investigation. And who should know your processes better than your experts? Substantial contribution from your process and samples experts is vital for the success of our cooperation.

Together we obtain information from your process data!

Our philosophy

Our philosophy is to establish long-term partnerships with leading companies. The preconditions for a common success are good team work and communication as well as the sharing of know-how and information. We are fully aware of the fact that we touch the core knowledge of our customers, therefore we spread our activities in different industries to avoid conflicts of interests in the cooperation with customers and partners.

Our vision

Our vision is to build up a Centre for Process Control with the focus on knowledge and experience in multivariate data analysis and process control which will offer innovative customers the possibility to integrate themselves to intensify the partnership in the project work, services, training and consulting.




To further enlarge the knowledge base and the capacity, cooperation is planned with partners from a well-established academic network and other companies in the field. The Centre will stimulate the exchange of knowledge and experience between all participants in compliance with security and protection of intellectual properties. Furthermore, the Centre will represent a platform for extended research and scientific activities as well as for additional services like instrument sharing and exchange of expertise in other fields.