Research area

Data modelling

A food can be affected by very many different factors, such as quality of raw materials, the production process, the instruments used during production and quality control. Getting a product to taste the same every time is not as straightforward as you might think.

The quality of biological raw materials varies widely because of seasonality, climate, differences between varieties and so on. When these raw materials are used in the food industry, this can often lead to the quality of the finished product also becoming unstable. Customers have clear expectations about a product's taste and consistency, and it is important for the industry to meet these expectations. Controlling production by means of statistics and data modelling helps to stabilise product quality.

Involved in many areas of research

Data modelling is used to perform scientific research throughout the entire process, from planning, through execution and statistical analysis to presentation. It is Nofima Mat's goal to be at the forefront of the development of statistical and mathematical methods, so that we can exploit the potential created by new and more powerful measurement techniques.

We have a long tradition of data modelling and statistics here at Nofima Mat and they have been in a state of continual development. Ever since the 1970s, Nofima Mat has been combining chemical ideas with statistical and mathematical calculation techniques, which have contributed to establishing companies and creating software programmes. This is a field that provides the necessary research basis upon which other areas of research depend, even though this may not occur to the customer.

Multivariate analyses

Because of the increased use of instrumentation and data tools there has been a huge increase in the volume of measured data. So it is necessary to find methods which can distil the essence out of all these numbers, enabling us to concentrate on what is important. Multivariate data analyses are methods for converting huge quantities of data from instruments, experiments and processes into information that can be interpreted. By using designed research, we are able to get the greatest possible amount of information out of a limited number of tests. This is quick, efficient and cost saving.

The methods thus developed can be used for:

  • better utilisation of raw materials
  • reducing wastage
  • reducing variations in the production process
  • more environmentally friendly production
  • efficient product development of, for example, food for better health

 

None Copyright: PhotoDisc

Relevant news

  • Investigating consumers’ eating habits

    18. March 2010

    New methods of data collection and analysis are giving the food industry a better insight into what consumers want. This could help more new products to survive in the marketplace.

Contact

  • Achim Kohler

    Research Scientist

    Phone: +47 64970240

    Cellphone: +47 901 80 765

Person

 

Research area within Data modelling

Data analysis in modern biology

This is an area of research in which data analysis is used to study biological mechanisms in relation to genes and environment. Research in this area helps us to understand the relationship between the quality of a food product and the genes of the animal or raw material. Research involves investigating the raw material's genes and their effect on the end product (grain, meat, milk etc.).

Process management

Before you can achieve an end product of consistent quality, you must have input knowledge about raw materials, processes, production, spectroscopy, chemometry and statistics. Process management provides rapid answers and more efficient production. There are often enormous quantities of data involved, that must be collated and made understandable for non-statisticians.

Sensometry

A sensometric map.

Sensometry is the connecting together of sensory perception, mathematics and statistical methods. The field embraces both analytical and measurement tools for sight, smell, touch, taste and hearing. Using statistical methods for analysing the data from sensory tests gives us the opportunity to make use of the results in both product development and quality control.