With regard to the climate change as well as the shortage and increase in price of natural resources, efficient resource management becomes more and more important, also for production processes. More than half of the sheet metal used in the automotive industry is wasted. Considering the amount of energy and the added value associated to the discards, enormous savings are possible. The approach of this research work is to avoid waste during a forming process by means of a continuous process monitoring and a self-learning intelligent tool interacting with the servo press. The flexibility of the servo press, which is so far used to optimize the single press stroke, only, shall be exerted to respond on altering material properties of blank sheets and the thereby changing forming behaviour. The sample process includes a two-tiered bending process, in which a conventional punch performs the pre-bending in the first step and an independent servo axis the compensation-bending in the second step. If an inadmissible bending angle is measured after the second step, the servo press stops automatically and the additional servo axis performs a teaching-stroke to adjust the compensation angle. After that, the continuous process starts again. This paper presents the springback compensation strategy which uses flexible correction algorithms based on data of the on-line process monitoring as well as data of the previous bending steps. Finite Element Analyses are carried out to validate the strategy.