Proceedings of the 2014 Industrial and Systems Engineering Research Conference Y. Guan and H. Liao, eds. Automated Process Planning and CNC-Code Generation Sean P. Turley a , David M. Diederich a , Bhanu K. Jayanthi b , Anuj Datar b , Christopher B. Ligetti a , Daniel A. Finke a , Christopher Saldana b , Sanjay Joshi b a Applied Research Laboratory, The Pennsylvania State University at University Park, Pennsylvania, 16802 b Department of Industrial and Manufacturing Engineering, The Pennsylvania State University at University Park, Pennsylvania, 16802 Abstract Traditional process planning systems require human input to interface between CAD and CAM systems. The purpose of the present study is the development of a generative process planning methodology for automated generation of CNC codes for 3-axis milling centers with the native CAD geometry as the only primary input. The proposed methodology uses geometric reasoning algorithms for recognizing machinable volumes, set up planning, fixture planning, tool selection, and tool path selection to generate machine codes. Rather than recognizing arbitrary defined machined features, specific features are recognized and mapped to machine tool paths in tool path planning software, thus allowing integration with commercially available software for CNC code generation. The geometric reasoning algorithms use boundary representation (B-rep) to decompose removal volumes into discrete milling operations. The algorithms presented here encompasses planar and cylindrical faces, but are flexible in that they can be extended to handle more complex geometry. Mastercam is then used (via an application programming interface) to generate tool paths using scripted point-by-point specification of tool paths, machine linking parameters and tool settings for each of these milling operations. Results pertaining to of end-to-end integration of the algorithms in the automated process planning and machining of test parts through multiple setups are presented and discussed. Keywords Automated Process Planning; CAD/CAM, CAPP, Automated CNC-code generation 1. Introduction Conventional manufacturing using Computer Numerically Controlled (CNC) machines from design inputs requires significant effort and time. Manufacturers have to understand and interpret Computer Aided Design (CAD) drawing inputs, map these to machinable features, determine setup directions, generate process plans, select appropriate tools and machining parameters. This data is fed to Computer Aided Manufacturing (CAM) systems for generation of CNC code. With increase in global competition in manufacturing, there is a need to automate this process to reduce the lead time, improve quality, productivity and product development life cycles. In this regard, Computer Aided Process Planning (CAPP) was introduced as a bridge between CAD and CAM systems that were developed independently. CAPP refers to activities done either automatically or with minimum human intervention to convert component designs into manufacturing instructions that describe how components are to be produced [1]. In the last three decades, many CAPP systems were developed and the literature in this domain is quite extensive. Xu et al. [2] classified CAPP systems depending on whether they are feature-based, knowledge-based, agent-based, or whether they use neural networks, genetic algorithms, fuzzy logic or Petri Nets. Among these, feature-based technologies have long been used in CAD/CAM integrations, as most CAM systems identify features as input data. Babic et al. [3] classified feature-based technologies into three types; design by features (DBF), interactive form feature recognition and automated feature recognition (AFR). Parts are designed using features from libraries in the DBF methodology. In interactive form feature recognition systems, the user determines recognition parameters from selected form features set, and using those instructions performs automatic recognition of features from CAD data. With an increase in the complexity of designs, it becomes difficult to create parts with features. Hence, DBF and interactive form feature recognition based technologies may not be applicable in most CAPP systems. AFR systems or simply feature recognition systems are systems where manufacturing features are identified from low-level design data, like boundary representation (B-rep) using different algorithms. The algorithms for these feature recognition systems were classified by Shah et al. [4] according to applied techniques: graph-based, volume