The general objective is to gain insight into the impact of the multistage production process of frozen par-fried potato fries on the quality of consumable Belgian fries and water management requirements, as a function of incoming potato’s variable characteristics. This insight will be provided through thorough investigation of the physicochemical impact of each industrial processing step and parameters applied, in combination with research into the development of highly efficient experimental plans. This approach will (1) allow to get fundamental insight in the combined role of process parameters and raw product characteristics on the quality of the consumable Belgian fry, and (2) provide a generic, efficient tailor design of experiments strategy for optimizing typical complex multistage and raw-material lot dependent production processes in agrifood.
The following sub-objectives are defined:
1. Perform strategic research into the creation of a highly informative, cost-efficient, generic experimental plan strategy for optimizing multistage processes tailored to highly variable input material, which is a typical situation for the agri-food sector. This sub-objective will allow to define which raw product x process parameter tests to perform to gain maximal insight with minimal experimental effort.
2. To gain mechanistic insight into the combined impact of different processing steps applied on a pilot scale line on the process water management and on the molecular dynamics within processed potatoes and their effects on the quality of the consumable Belgian fry. This will be achieved by applying the novel, generic experimental plan strategy to the specific case of the production process of par-fried frozen potato fries. This sub-objective will produce the required experimental data to gain
insight with respect to the combined role of process parameters and raw product on the quality of the consumable Belgian fry. In addition, this sub-objective will enable to reach the modelling and optimization goals, specified further (see 3, respectively 4 and 5).
3. Create an optimization framework that allows to find the best possible process settings, generic for any particular lot of raw potatoes to be processed. Apply this generic framework to the experimental results from 2 and implement the obtained framework in a software code that can be used as a pilot expert system, replacing the traditional trial-and-error approach. This sub-objective will produce a statistical model and related optimization procedure allowing to define optimal process settings for any
given batch of raw potatoes.
4. To evaluate and validate the pilot expert system by aiming to reach specific optimization goals for respectively selected characteristics of the consumable Belgian fry (i.e., acrylamide level (ALARA level), fat content (25% reduction), texture (25% improvement), grey discoloration (reduction in pyrophosphate by 50%, while limiting increase in discoloration) or the process water management for three lots of incoming potatoes. This sub-objective will verify the validity of the pilot expert system by
considering each of the selected outcomes with respect to product quality and water management separately (single output optimization).
5. To generate a proof-of-concept that the pilot expert system is able to optimize the multistage pilot scale production process of frozen par-fried potato fries with respect to the whole set of considered quality characteristics of the consumable Belgian fry and the process water management, using three lots of incoming potatoes. This sub-objective will verify the validity of the pilot expert system by considering the combined outcomes with respect to product quality and water management together (multiple output optimization).