The increasing complexity of the manufacturing industry demands computer numerical control (CNC) machining, a massively used process that is equipped with an intelligent feature. Not only does its control function revitalize competition in a sustainable manufacturing ecosystem but also advances the Industry 4.0 readiness. This paper develops an intelligent computer-aided process planning (i-CAPP) based on two widely accepted performance measures: manufacturability and efficiency. Machine tapping is a very popular CNC machining process making threaded holes, from which there is an investigation of two manufacturability scenarios. The first scenario deals with all holes with the same size that undergo the same operation (OP). The second one handles the groups of holes with different sizes machined by various OPs. Meanwhile, the study of efficiency focuses on point-to-point drilling, tapping, and chamfering. This research work verifies such indicators by proposing hybrid-two-stage optimization algorithms: 1) traveling salesman problem and 2) innovative Tabu Search. The i-CAPP verification of both indicators was confirmed in an application with a five-axis CNC tapping machine to realize the Industry 4.0 readiness. Extensions of the research in the pursuit of higher intelligent manufacturing platforms are discussed.

In recent years, the use of reinforcement learning and imitation learning to complete robot control tasks have become more popular. Demonstration and learning by experts have always been the goal of researchers. However, the lack of action data has been a significant limitation to learning by human demonstration. We propose an architecture based on a new 3D keypoint tracking model and generative adversarial imitation learning to learn from expert demonstrations. We used 3D keypoint tracking to make up for the lack of action data in simple images and then used image-to-image conversion to convert human hand demonstrations into robot images, which enabled subsequent generative adversarial imitation learning to learn smoothly. The estimation time of the 3D keypoint tracking model and the calculation time of the subsequent optimization algorithm was 30 ms. The coordinate errors of the model projected to the real 3D key point under correct detection were all within 1.8 cm. The tracking of key points did not require any sensors on the body; the operator did not need vision-related knowledge to correct the accuracy of the camera. By merely setting up a generic depth camera to track the mapping changes of key points after behavior clone training, the robot could learn human tasks by watching, including picking and placing an object and pouring water. We used pybullet to build an experimental environment to confirm our concept of the simplest behavioral cloning imitation to attest the success of the learning. The effectiveness of the proposed method was accomplished by a satisfactory performance requiring a sample efficiency of 20 sets for pick and place and 30 sets for pouring water.

This paper describes the fabrication of a series of micro ball-ended stylus tips by applying micro-EDM (Electrical Discharge Machining) and OPED (One Pulse Electrical Discharge) processes, followed by a manual assembly process of a static tri-switches tactile structure on a micro-CMM (Coordinate Measuring Machine). This paper further proves that the essential performance of the proposed system meets an acceptable benchmark among peer micro-CMM systems with a low cost. The system also adjusts for ambient temperature and humidity as the ordinary lab environmental conditions. For demonstration, several experiments used a randomly selected glass stylus with the diameters of stem and sphere of 0.07 mm and 0.12 mm, respectively. By leveraging research guidelines and common practice, this paper further investigates the probing relationship between measurement accuracy and its associated critical characteristics, namely triggering scenarios and geometric feature probing validation. The experimental results show that repeated detections in the uncertainty, in vertical and horizontal directions of the same point, achieved as small as 0.11 μm and 0.29 μm, respectively. This customized tri-switches tactile probing structure was also capable of measuring geometric features of micro-components, such as the inner profile and depth of a micro-hole. Finally, extensions of the proposed approach to pursue higher accuracy measurement are discussed.