The logistical control and order release of structurally dynamic systems represent a great challenge. In contrast to classical shop floor production, these systems can involve transitions between the two different organizational forms job shop production and flow production. This results in a system that is susceptible to turbulence and malfunctions. This paper presents an approach that improves the robustness of these systems and achieves a stable system state even under external influences. Decentralized stock control loops control the WIP of the individual stations in a closed loop. When organizational transitions occur, interlinking losses arise, which reduce the productivity of the system. Similar effects can cause disturbances and blockages. The introduction of a robustness controller adapts the input parameters of the order release to the current system status, so that an adapted reaction takes place in the event of strong disturbances.