OATAO is an open access repository that collects the work of Toulouse researchers and makes it freely available over the web where possible.

The management of end-of-life systems is becoming a major concern for systems manufacturers as the negative impact of these systems on the environment is a matter of increasing public awareness and their appropriate treatment offers economic opportunities. In this context, the disassembly of these systems in order to recycle their components is a possible and sound option that can make it possible to sustain economical progress while respecting environment requirements. The work undertaken in this paper considers modelling and optimizing issues of such disassembly activities. An integrated approach is proposed to model and optimize the selection of valuable components of end-of-life systems, their recycling options and the way to obtain them. Because the framework of such problems is highly uncertain, we propose the use of Bayesian networks and their extension in terms of influence diagrams as mathematical tools for structuring and managing uncertainties. This approach allows taking into account uncertainties rising from different sources on one hand and as a support for optimization on the other hand.

The management of end-of-life systems becomes more and more important due to the awareness of their environmental impact. In this context, the disassembly process requires more attention with the ultimate goal to make profit. In this paper, we propose a new approach to determine optimal disassembly plan of an end-of-life system by using bayesian network. To take advantage of some existing approaches that use Petri Net to model such process, a Petri Net model is first established and then translated to Bayesian Network in order to take into account inevitable uncertainties associated to such process.