A novel algorithm for optimal buffer allocation in automated asynchronous unreliable lines

The Buffer Allocation Problem is a well-known optimization problem aiming at determining the optimal buffer sizes in a manufacturing system composed by various machines decoupled by buffers. This problem still has scientific relevance because of problem complexity and trade-off between conflicting goals. Moreover, it assumes industrial relevance in reconfigurable manufacturing lines, where buffer sizes can be easily adapted to the production scenario. This work proposes a novel algorithm integrating performance evaluation and optimization by means of throughput cuts based on a linear approximation. Numerical results show the validity of the proposed approach with respect to the traditional gradient-based method. Moreover, an industrial case study integrating the proposed approach into a decision-support system for the buffer allocation and reallocation is analyzed.