Integrating Cloud Computing, Bayesian Optimization, and Neural-Additive Modeling for Enhanced CAM Systems in 5-Axis Milling

This publication introduces the development and application of an advanced CAD/CAM/CAE system that leverages the computational capabilities of an edge-cloud infrastructure. The developed system consists of containerized technology models for various complex process planning simulation tasks in 5-axis-milling, such as toolpath calculation, cutter-workpiece-engagement, uncut chip geometry or cutting force simulations. Moreover, a Bayesian optimization (BO) algorithm is coupled with the system enabling multi-objective optimizations of the considered machining operations by varying predefined CAM-parameters. The result of the optimization consists of a set of Pareto-efficient solutions. Each solution realizes a different tradeoff between the technological objectives of the process planner. Since mastering the complexity of the design space is a major challenge in today’s CAM programming workflow, a Neural-Additive Model (NAM) is coupled with the system improving guided search through the configuration/result space. This reduces the convergence time to an optimal CAM parameter set. The coupling of the cloud computing, multi-objective optimization and artificial intelligence method with a CAM-kernel is demonstrated based on the process design of 5-axis milling operations for a blade-integrated disk (blisk).