Search

Agent-based modeling and simulation is a method well suited for studying individual behavior and interactions among members of a population connected by social networks. Although the development of such simulation models can be relatively complex, it is even more challenging to develop models that are empirically valid. In the case of reverse logistics, the sophisticated and difficult-to-predict behavior of consumers must be modeled. In this paper, an agent-based simulation model of consumer behavior and interactions was configured to conduct a case study of the voluntary deposit collection program for wine bottles in the Val-Saint-François region of Quebec. As this collection program was officially launched in 2019, two empirical samples were obtained to test the validity of the model and study how social interactions such as word of mouth contributes to the success of the collection program. The first sample represents the amount of glass collected during the last 26 weeks of 2019, while the second sample covers the first 13 weeks of 2020. Having observed an increase in collection rates between 2019 and 2020, word of mouth was introduced into the model to explain this phenomenon. Statistical tests show that the model is indeed valid with the inclusion of diffusion of awareness, as the simulation results are significantly consistent with the empirical data. The validation of the model demonstrates the viability of using multiple heterogeneous data-sources to configure a simulation model based on the Theory of Planned Behavior without using Structural Equation Modeling.

Nous réfléchissons au rôle central de la confiance dans le processus de déploiement de la R-I (recherche-intervention). Nous envisageons la confiance comme un facteur décisif de performance institutionnelle (Thiébault, 2003), dont le rôle est particulièrement mobilisé en période de changement, que la RI se propose d’accompagner contractuellement (Marcel & Bedin, 2018). À partir d’une comparaison entre deux sphères socioculturelles, la France et l’Amérique du Nord, nous mettons en exergue l’importance de situer, et d’évaluer la confiance en tant que processus et paramètre fondateur de toute recherche participative. À titre d’exemples illustratifs, nous mobilisons les travaux développés au sein de l’UMR-EFTS.

Nous poursuivons nos réflexions sur la nature de la confiance et de son rôle central au sein des démarches de R-I (recherche-intervention), inscrites dans une perspective historico-culturelle. Décrivant le salariat en France tel qu’il s’est développé pour devenir le principal mode d’employabilité au sein des milieux professionnels de l’éducation, nous discutons des concepts de vulnérabilité et de défiance dans la R-I, source d’incompréhension et de conflit potentiel. Nous proposons certaines clés de compréhension et de réflexion, pour finalement avancer qu’il est fécond pour la·le chercheur·se d’identifier et canaliser la défiance (et la méfiance) afin de favoriser le déploiement d’un objectif majeur de la R-I : l’émancipation.

The authors wish to make the following corrections to this paper [...]

This paper presents an agent-based simulation model for end-of-life product flow analysis in recuperation and recycling supply networks that focuses on individual consumer behaviors. The simulation model is applied to a deposit-return program on wine bottles that could be developed in the province of Quebec. Canadian data was used to calibrate and validate the model. A series of experiments was then conducted with three artificial populations to analyse how they would react to several implementation scenarios of this end-of-life product flow strategy. The results suggest that the distance to the nearest depot is an important decision factor, but less predominant than the ownership of a private vehicle and the deposit value. The results also indicate that the use of agent-based modeling combined with the theory of planned behavior (TPB) can produce modular behavior models, that are intuitive and simple, to better understand consumer-behavior-driven supply chains. Such models can be used to give insights to decision-makers and policy-makers about the potential performance of end-of-life product flows strategies and further facilitate efficient resource management.