Computational intelligence in food safety:

Computational intelligence is a term that refers to the use of software tools to extract information and build knowledge representations for complex systems such as a food production and distribution chain.  We use the tools of computational intelligence to help humans (e.g. food processors and policy makers) consider a wide range of factors as they make decisions related to food safety.  

Current Projects

Multi-factorial framework for risk prioritization of foodborne pathogens:

We have developed a knowledge-based system that provides measures for the analysis and prioritization of microbial risks in food systems based on 4 major dimensions of risk: public health impact, consumers’ perception and acceptance of risks, market impacts and societal values.  We have extracted information from health data bases and experts’ opinions to better estimate risk measures for the Canadian food system.  We are also exploring the use of multi-criteria decision methods to rank microbial hazards based on aggregation of multiple risk factors. 
This work is done in collaboration with researchers from Agricultural Economics (Caswell, Cranfield, Henson, Anders, Schmidt), Public Health Agency of Canada (Fazil) and Health Canada (Farber).

Ruzante, J.M., Davidson, V.J., Caswell, J.A., Fazil,A., Cranfield, J.A.L., Henson, S.J., Anders, S.M., Schmidt, C. and J. Farber 2009  A Multi-Factorial Risk Prioritization Framework for Food-borne Pathogens, Risk Analysis (doi=10.1111_j.1539-6924.2009.01278).

Process Risk Model to Estimate Incidence:

A system-based analysis is used to model hazard levels through a food processing and distribution system (including consumer handling) and to estimate total illness in specific populations due to consumption of infected products.  The process model must incorporate key processing stages that affect microbial growth/death and potential cross-contamination in order to estimate hazard levels and to evaluate the impact of interventions.  

Past Research Projects

Multi-criteria Decision Analysis (MCDA) Methods:
Daza Donoso used a number of MCDA methods to compare and rank risks posed by different pathogen-food combinations. 
Daza Donoso, C.A. 2008. Application of Multicriteria Decision Analysis Tools to the Prioritization of Microbial Hazards in Food Systems, M.Sc. thesis, University of Guelph.

Computational Intelligence in Process Control:
Drying is a common food processing operation that has substantial impact on food product quality (e.g. biological activity, rehydration characteristics). Control strategies must consider quality as well as economic factors such as productivity and energy costs. Our application to ginseng drying is relevant to food processing in Ontario but the general strategy can be extended to a number of other applications. The objective is to design a drying control system that preserves quality at user-defined levels and also achieves the highest moisture removal rates, given quality set points. The control system has been developed and tested on laboratory- and pilot-scale equipment.


Dr. Valerie Davidson, P.Eng.

519-824-4120 Ext. 54367