This paper proposes a decentralised planning optimisation approach by using mathematical programming and negotiation mechanisms in a collaborative network. Two partners addressing the replenishment and production stages are considered. Concretely, production scheduling plans and material requirement plans, modelled by two mixed-integer programming models, are modified according to the established negotiation rules between the two collaborative partners. The main contribution of this paper is the improvement of the replenishment and production plans upstream of the network; proposing a decentralised collaborative planning process. The validation of the proposal is done using data based on an automotive industry.