Agri-food sector performance strongly impacts global economy, which means that developing optimisation models to support the decision-making process in agrifood supply chains (AFSC) is necessary. These models should contemplate AFSC's inherent characteristics and sources of uncertainty to provide applicable and accurate solutions. To the best of our knowledge, there are no conceptual frameworks available to design AFSC through mathematical programming modelling while considering their inherent characteristics and sources of uncertainty, nor any there literature reviews that address such characteristics and uncertainty sources in existing AFSC design models. This paper aims to fill these gaps in the literature by proposing such a conceptual framework and state of the art. The framework can be used as a guide tool for both developing and analysing models based on mathematical programming to design AFSC. The implementation of the framework into the state of the art validates its. Finally, some literature gaps and future research lines were identified.

No abstract

No abstract

Agri-food supply chains are subjected to many sources of uncertainty. If these uncertainties are not managed properly, they can have a negative impact on the agri-food supply chain (AFSC) performance, its customers, and the environment. In this sense, collaboration is proposed as a possible solution to reduce it. For that, a conceptual framework (CF) for managing uncertainty in a collaborative context is proposed. In this context, this paper seeks to answer the following research questions: What are the existing uncertainty sources in the AFSCs? Can collaboration be used to reduce the uncertainty of AFSCs? Which elements can integrate a CF for managing uncertainty in a collaborative AFSC? The CF proposal is applied to the weather source of uncertainty in order to show its applicability.

Some small farms are forced to waste a part of their harvests for not reaching the quality standards fixed by consumers. Meanwhile, modern retailers (MR) are interested in selling more quality products to increase their profits. MR could invest in a collaboration program so the small farmers could have access to better technologies and formation to increase the proportion of quality products. Unfortunately, the demand, the quantity of harvest, the proportion of harvest being of quality, and its increase with each investment are uncertain parameters. A fuzzy model considering these uncertainties is proposed to determine the investments that MR should made to maximize the profits of the supply chain in a collaboration context. A method to transform the fuzzy model into an equivalent crisp model and an interactive resolution method are applied.

Open Archive Toulouse Archive OuverteOATAO is an open access repository that collects the work of Toulouse researchers and makes it freely available over the web where possible To cite this version: ZaratéAbstract. Decision making for farms is a complex task. Farmers have to fix the price of their production but several parameters have to be taken into account: harvesting, seeds, ground, season etc… This task is even more difficult when a group of farmers must make the decision. Generally, optimization models support the farmers to find no dominated solutions, but the problem remains difficult if they have to agree on one solution. In order to support the farmers for this complex decision we combine two approaches. We firstly generate a set of no dominated solutions thanks to a centralized optimization model. Based on this set of solution we then used a Group Decision Support System called GRUS for choosing the best solution for the group of farmers. The combined approach allows us to determine the best solution for the group in a consensual way. This combination of approaches is very innovative for the Agriculture domain.

No abstract

The objective of this paper is to examine the use and applications of reinforcement learning (RL) techniques in the production planning and control (PPC) field addressing the following PPC areas: facility resource planning, capacity planning, purchase and supply management, production scheduling and inventory management. The main RL characteristics, such as method, context, states, actions, reward and highlights, were analysed. The considered number of agents, applications and RL software tools, specifically, programming language, platforms, application programming interfaces and RL frameworks, among others, were identified, and 181 articles were sreviewed. The results showed that RL was applied mainly to production scheduling problems, followed by purchase and supply management. The most revised RL algorithms were model-free and single-agent and were applied to simplified PPC environments. Nevertheless, their results seem to be promising compared to traditional mathematical programming and heuristics/metaheuristics solution methods, and even more so when they incorporate uncertainty or non-linear properties. Finally, RL value-based approaches are the most widely used, specifically Q-learning and its variants and for deep RL, deep Q-networks. In recent years however, the most widely used approach has been the actor-critic method, such as the advantage actor critic, proximal policy optimisation, deep deterministic policy gradient and trust region policy optimisation.