Artisanal and small-scale gold mining in Burkina Faso: Suggestion of multi-agent methodology as a complementary support in elaborating a policy

In order to better manage artisanal and small-scale gold mining, the Burkinabe authorities have planned to build a suitable methodological support as an aid to elaborating appropriate policies and actions. The developed methodology concerns (a) at a spatial level, the generation and analysis of a geological resources map needed by the artisanal miners and (b) at a socioeconomic level, analysis of the miners' activity. This paper reports and discusses the results of our suggestion to also introduce the use of an approach known as multi-agent system (MAS) as a complementary part of this initial methodology, at a downstage level to the above two analysis stages. MAS methodology could be used to develop simulation models to forecast the future of the activity. MAS is appropriate for describing the dynamics of systems where it is impossible to obtain all output data of a model from a purely mathematical or statistical transformation of input data. Our suggestion involves (i) demonstrating the capacity of the approach to simulate all parameters needed by mining policy makers and (ii) evaluating to what extent the methodology is accepted by them, as well as, at a more scientific level, the literature regarding mining system simulation.Furthermore, although the core of this paper focuses on the Burkina Faso issue, an extended discussion is introduced at the end of the paper regarding the possible interest of applying the MAS approach to the industrial mining sector. The idea, in particular, is to simulate scenarios of supply shortages of critical industrial raw materials globally and in national markets.As a result of this work, a demonstration has been made that the models developed by the MAS approach can reproduce and simulate players' activity in both mining types. However, convincing the mining sector to integrate MAS as a full part of a methodological support for policy makers remains a challenge. Explanations for this state of affairs and proposals for progress regarding MAS acceptance are provided.

► We want to show the interest of the multi-agent modelling method in the mining sector.
► We then develop application models of artisanal gold mining in Burkina Faso.
► As an extension, we also develop a market model of industrial critical materials.
► MAS is an interesting approach, as a complement to a classical mathematical approach.
► However, the acceptance of the method by mining policy makers remains a challenge.