The design of a compact laser scanner and an integrated simulation environment for smart manufacturing

Abstract

The development in material science and robotics makes 3D printing large building structures possible and desirable thanks to its flexibility and efficiency. Automated 3D printing of building structures has many benefits ranging from reducing construction cost to helping make habitats for space explorations. However, most of the existing robotic 3D printing solutions require human supervision for monitoring and intervention. We believe this lack of autonomy is partly because of the lack of a suitable sensor that could provide the necessary feedback information, and the difficulty in developing control algorithms on physical hardware that is expensive and dangerous to operate.

In this work, we present a design framework for a miniaturized laser scanner that could be used for in-situ inspection in additive processes by providing submillimeter-accurate, real-time depth reconstruction of the printed material and imagery feedback. We also present a novel, all-in-one simulation environment for accelerating control software development based on the simulated sensory feedback. The proposed simulation environment is equipped with photorealistic rendering, sensor simulation, additive fluid material simulation, robot arm and robotics software integration.

We show the precision of our laser scanner by comparing the measured widths of some 3D printed gaps against the ground truth. To show the effectiveness of our robotic manufacturing simulation environment, we modeled a full robotic additive printing process with our laser scanner, an additive thermal plastic extrusion nozzle and a servoing robot arm. We then showed a simulated in-situ scan result using a simple sensor placement strategy that maximizes scan coverage given our sensing constraints.