Manufacturing methods for aerospace composite parts are various. Vacuum assisted resin transfer moulding (VARTM) is one of the common methods. It requires manual draping of fabrics to produce preforms. The cost and quality of this method depend on the draping strategy, which is the sequence of draping operations performed by staff as they lay fabrics on moulds. These sequences can be very numerous with many possible starting points, yarn orientations, and fabrics used. In this group project, a predictive software tool is designed ultimately to identify the best draping strategy to reduce cost and improve quality in producing preforms. Generation of preform flat patterns for the CNC cutting tables, and of plybooks for providing instructions to shop floor staff, are developed and integrated into the predictive tool so that the manual draping process can potentially be conducted under the best draping strategy. The work presented in this thesis details the solutions for finding flat patterns for the best draping strategy for the CNC cutting tables, and the generation of plybooks for providing instructions to shop floor staff. The work is organized into four main parts: developing flattening algorithms for three types of base surfaces (BaS) defined in this project, developing a subsequent stitching algorithm aimed at stitching individual flattened BaS into a unified flat pattern for a given mould, lab validations to assess the proposed flattening and stitching algorithms, and establishing the capability to generate instruction .DXF files and plybooks for any received draping strategy. These algorithms are integrated into the predictive tool to facilitate manual draping processes under the optimal draping strategy.