Supply chain design with an environmental performance-sensitive demand – CONCLuDE

- Identification of criteria for the sensitivity of demand to environmental performance. Based on an in-depth study of the literature and interviews with decision-makers, we understand how to measure this attribute.
- Mathematical characterization of demand. Modeling the sensitivity of demand to environmental performance presents a number of challenges, as this sensitivity depends on many factors, such as the sector concerned, market competition, and customer behavior (for example, the literature is not unanimous on the willingness of customers to seek out products with good environmental characteristics, especially if this leads to a reduction in environmental impact).
good environmental characteristics, especially if this leads to higher prices). Based on the first step, we carried out an empirical study of the agri-food sector in France to characterize demand, which we translated into mathematical functions.
- Mathematical modeling of CL design and management problems with environmentally sensitive demand and model resolution. We have developed two types of models: (1) Linear integer programming (LINP) models, which offer modeling flexibility for the integration of different logistical decisions, but make it difficult to take complex demand modeling into account. We have developed these models for CL design in the agri-food context. (2) Analytical models, which allow complex demand modeling to be considered, but generally propose a simple modeling of CL decisions and structure. We have developed these models to understand the impact of endogenous demand modeling and the sensitivity of demand to environmental performance and selling price, initially for a generic case, and then for the retail sector, which generally has a simple LC structure and a limited number of decisions compared with the industrial sector.
- Model validation and exploitation. The challenge posed by the PLNE models lies in the difficulty of generalizing the behaviors observed on a few instances and drawing generic recommendations from them. To achieve this, we carried out a large number of experiments. For the analytical models, we characterized the optimum whatever the data instance considered (under certain conditions). We analyzed the results and the behavior of the optimal solutions and profits to derive managerial recommendations.

The project shows, for the French agri-food sector, that customer demand does not increase linearly with the environmental performance of products, and that above a certain level of environmental performance, an improvement in this performance has no impact on demand. With regard to LC management, the project team shows that in the presence of customers' environmental awareness, local (short) LCs can be more profitable than international (long) LCs, and that better environmental performance can be achieved without deteriorating economic performance. Building on the models developed during the project, the project team has extended the scope of the work to study sectors where demand is sensitive to delivery time and price, in particular the e-commerce sector.

The project has enabled us to understand the impact of customers' environmental awareness on product demand and company logistics decisions. A key finding is that customers' environmental awareness can lead companies to offer more environmentally-friendly products without jeopardizing their economic performance. We have studied the agri-food and retail sectors. Our results can be generalized to other sectors. Last but not least, the project has enabled us to identify interesting research issues in the agri-food sector, such as the impact of climate uncertainties on companies' environmental decisions, topics which we are currently developing and which may give rise to new research projects.

Hammami, R.; Asgari, E.; Frein, Y.; Nouira, I. Time- and price-based product differentiation in hybrid distribution with stockout-based substitution. European Journal of Operational Research. 2022, 300 (3), 884-901.

Asgari, E.; Frein, Y.; Hammami, R. On Retailers’ Profit in A Price- and Greenness Sensitive Market with Different Demand Functions. Conférence Internationale Génie Industriel QUALITA. Grenoble, France. May 2021.

Asgari, E.; Hammami, R.; Frein, Y. Retailing Competition for Substitutable Products in Greenness- and Price- Dependent Market. International Conference in Information, Logistics and Supply Chain. Texas, United States. Apr 2020.

Nouira, I.; Gondran, N.; Hammami, R. Organisation de la session spéciale “Supply chains of the future: Towards greener and more local supply chains” dans la Conférence Internationale Génie Industriel QUALITA. Grenoble, France. Mai 2021.

This project aims to revisit the Supply Chain (SC) design models while considering an endogenous demand sensitive to the environmental performance, which distinguishes our work from existing models where demand is usually modeled as an exogenous parameter that does not depend on the model decisions. Indeed, on the one hand, the environmental performance has become a factor of competitiveness for companies and an important purchasing criterion for customers whether in the context of B-to-B or B-to-C. On the other hand, the environmental performance depends on the decisions undertaken at the design of SC such as location/allocation of production sites, choice of suppliers, selection of manufacturing technologies and transportation modes, etc.

While this project is theoretical research-oriented, we are aware of the necessity of validating our assumptions and models by considering real-worlds situations. Two industrial areas are particularly interesting for us: mechanical manufacturing and food industry. With this objective in mind, we will collaborate with “pôle LUTB”, “pôle Viaméca” and the company Diana Food that accepted to provide us with data and relevant examples.
First, we have to determine the attributes that affect demand. For instance, the carbon footprint is often considered, but it is a shortcut that is increasingly challenged by some comprehensive approaches such as the Life Cycle Analysis. Then, we have to build mathematical equations establishing the relationships between the environmental performance of a product and its demand level. The next step is to integrate the demand functions established in the previous phases in SC design models and to solve these models. Two main categories of models will be considered:

- mixed integer linear programming that can integrate several decisions simultaneously, but often do not allow characterizing analytically the optimal solution,

- analytical models that are more generic but offer the possibility of integrating complex demand functions and obtaining analytically the optimal solution.

We will use advanced operations research techniques for modeling and solving. Finally, we will dedicate the last step to the validation and experimentation of the proposed models. In fact, we will use our models to show that ignoring the sensitivity of demand to the environmental performance could lead to inappropriate decisions. We will also try to derive insights such as the trade-off between local and international SC when the customers are sensitive to the environmental performance, the impact of customers’ environmental awareness on logistics decisions, the expected gain of a company that takes into account the sensitivity of customers to environmental performance and adapts its SC to meet customers’ requirements, etc.