Fully Automated Data Acquisition for Laser Production Cyber-Physical System

The many tunable parameters involved in laser processing, such as wavelength, pulse duration, pulse energy, and scan speed, not to mention various other complicating factors on the material side, makes it practically impossible to reliably find an optimized parameter set to realize a specific processing target. Currently, an acceptable parameter set is mainly found by tapping the experience and intuition of skilled people within the present production system. However, such methods do not scale to the mass-customization needs of the coming super-smart society, and it has become critical to develop ways to transfer such human experience and intuition to a more scalable setting: namely, the cyber-space. A major challenge in developing a cyber-space solution has been augmenting the limited experimental and theoretical insights of the laser processing phenomenon to the specific problems at hand. Here, we focus on automated data acquisition systems coupled with artificial intelligence (AI) methods to overcome this technological gap. We propose ways to realize cyber-physical systems specializing in specific facets of laser production by showing experimental results from four kinds of automated data acquisition systems. We lastly discuss such methods in context as an important first step to creating an AI based cyber-physical simulator.