The case studies given in the previous sections reflect well the present situation of AISF: AISF has a great potential to fill the existing gap in small-batch sheet metal forming of asymmetric shapes. AISF can yield excellent results for some parts, while for others, unforeseeable problems can arise. Despite almost 15 years of development, the process limits have not yet been generally overcome. Only with a deeper knowledge of the process mechanics, suitable process strategies and, most importantly, process models, AISF will become "ready-to-use" for industrial applications. In sum, the following developments seem necessary:. CAE: CAD-CAM tools are needed that are adapted to the requirements of AISF. . Process mechanics: The phenomena that are responsible for the "extended" forming limits and excessive springback are not yet fully understood. The same holds for wrinkling. Without that understanding, the feasibility of a given design cannot be assessed. . Process control: Infinitely many tool paths can be derived for a given part design. In principle, this flexibility offers the possibility to control the material flow in AISF. However, the material flow and its control are not yet understood. . Accuracy: Residual stresses in a part can yield unforeseeable consequences in the trimming stage such that the geometry created by the forming operation can be lost (cf. section 1.5). Residual stresses and springback are also found in deep-drawing. However, research in AISF is lagging behind the knowledge on springback control that has been acquired in deep-drawing over decades of research and development. Thus, it seems to be particularly important to improve the accuracy of the process. This is supported by a "search for applications" conducted by Allwood et al. [AKD05], who found that the range of possible applications of AISF depends mainly on its accuracy.This thesis aims at contributing to the points mentioned above. As a starting point, the stages that the part in the case study of section 1.5 ran through will be formalised to describe a consistent, general work flow for AISF. The stages to which this thesis contributes contain references to the corresponding chapters and sections.