• Reduce text

    Reduce text
  • Restore text size

    Restore text size
  • Increase the text

    Increase the text
  • Print


A decision support software for the packaging of fresh fruits and vegetables

The storage of fresh fruits and vegetables under an O2-depleted atmosphere using a packaging system whose gas permeability (O2 and CO2) adjusts perfectly to respiration of the product can increase their shelf-life.  Scientists have now developed a mathematical model that can guide the choice of packaging for a product according to its gas transfer parameters.

Marché de Copacabana ,à Rio de Janeiro. Etale de fruits et légumes tropicaux. Légumes préparés pour la vente. © BOSSENNEC Jean-Marie

The mathematical model that enables the modelling of gas transfers and associated respiration reactions in a packaging/product couple was considerably refined and then validated during the TAILORPACK project (2007-2011, www.tailorpack.com), funded by the French National Research Agency (ANR).  An online application is now accessible to all project partners and actors in the sector, via the Technical Centre for Fruits and Vegetables (CTIFL).To enrich this application and produce a true decision support tool (called DSS) for packaging choices, the researchers have been working in the context of the European EcoBioCap project (2011-14) to develop a novel tool for the design of biodegradable packaging systems using renewable resources and built by applying a knowledge engineering approach.  It is intended for use by the food packaging industry, the target application being the fruit and vegetable sector.

The Figure below summarises the design path adopted for this tool.  The aim of the method is to be able to draw up reasoned recommendations regarding the choice or design of a packaging system for a particular food product.  This can be achieved through multicriteria analysis of the preferences expressed by actors in the sector, while taking account of packaging characteristics recorded in a database.

description. © inra, inra
description © inra, inra

Synthetic representative of the knowledge integration approach

This work could be broken down into three phases:

  1. exploitation of data concerning the sector under study (characteristics of the product and packaging materials) using emergent techniques for data extraction arising from heterogeneous sources, so that they can be stored in the databases associated with DSS;
  2. modelling of the issues expressed by different sector actors (e.g. packaging manufacturer, user company in the food industry, consumer, etc.); identification of obstacles such as contradictory viewpoints; search for compromise;
  3. modelling of scenarios (e.g. choice of a biodegradable packaging material for "Charlotte" variety strawberries which is preferably transparent and costs less than €1.5 per kilo, etc.) by interrogating the databases.

A knowledge engineering method as an aid to decision-making

The most important recent findings concern more specifically the process for decision-making which, based on the "specifications" of the packaging sought (food to be packaged, shelf-life, storage temperature, etc.), and resulting from a consensus between actors in the industry, can generate a list of the most pertinent packaging materials.
More precisely, based on the specifications, and using a reverse engineering approach, the optimum technical characteristics of the packaging (oxygen and CO2 permeability) are determined by digital simulations performed using the Tailorpack mathematical models.  The materials selected using these technical criteria are then compared with the constraints (e.g. cost) and preferences (e.g. transparent, compostable materials, using local renewable resources, etc.) of the user of the tool so as to identify the packaging that best responds to a series of criteria that may sometimes be contradictory.
The decision-making process thus implemented is robust and novel, insofar as firstly, it has been designed to take account automatically of uncertainties regarding scientific knowledge and the reliability of the data available, and secondly because it has required the design of a novel method for multicriteria optimisation that interrogates databases so that they can generate the most appropriate packaging materials.

Which model can arbitrate in the packaging sector?

During this project, we have supposed that the "specifications" for the packaging sought result from a consensus between actors in the sector, including researchers and the developers of new materials.  However, the objectives of these different actors may diverge, so that decision-making requires the resolution of arbitration problems.
A scientific approach that is innovative regarding the analysis of such sectors, will be implemented under the EcoBioCap project and enable the introduction of explanatory factors into the decision-making process.  Several open questions are being studied: which type of representation, and which reasoning model can take account of the contradictory viewpoints in play?  How can the diverging priorities of different actors be reconciled?  How can a decision support system be designed that is capable of resolving conflicts and reaching a compromise?

An argument-based knowledge engineering method

The teams involved in the project intend to propose a generic, argument-based method to aggregate the preferences of actors in the sector.  This approach is innovative at both the theoretical and applicative levels.  Indeed, argument-based decision-making is a subject that has hitherto been little considered in the international literature, the first formal article only being published in 2009.  Nevertheless, this approach seems highly relevant to supporting decision-making in different sectors and appears to be capable of overcoming crucial obstacles linked to arbitration within sectors, such as an analysis of the stakes, the consequences of a decision and explaining a process that led to the decision.  The packaging sector is particularly appropriate to implementation of this approach because of its dynamism and innovativeness, and also due to the numerous safety, environmental, technical and economic constraints which prevail.

This work was carried out in the context of the ANR TaylorPack project.  It is continuing with the partners in the European EcoBioCap project.
Ecoefficient Biodegradable Composite Advanced Packaging.  http://www.ecobiocap.eu/

Scientific contact(s):

  • Patrice Buche (+33(0)4 99 61 22 83) UMR Ingénierie des Agropolymères et Technologies Emergentes
Associated Division(s):
Science for Food and Bioproduct Engineering
Associated Centre(s):