The advances in production planning and control in the last decades have confirmed the centralized nature of the planning function. However, centralization has disadvantages when quick response to local production problems is required. Therefore, renewed interest in decentralized approaches emerges. This paper investigates the possibility to use intelligent products for decentralized handling of disturbances. Intelligent products are aware of their local context and they can negotiate with local manufacturing resources. Therefore, local solutions for problems can be suggested, virtually at the same time at which the problem occurs. The paper demonstrates the viability of this approach via a simulation study. For reasons of comparison, the TAC SCM environment is used. Moreover, robustness is taken as an additional measurement of performance. The results of the simulations are encouraging.