3D graphic laser safety notification system for industrial human-robot cell

• Main Authors: Seet, Gerald Gim Lee, Viatcheslav, Iastrebov, Pang, W. Ching, Tjahjowidodo, Tegoeh
• Other Authors: School of Mechanical and Aerospace Engineering
• Format: Conference or Workshop Item
• Language: English
• Published: 2021
• Subjects: Engineering::Mechanical engineering::Robots; Human-robot Interaction; Autonomous Vehicle
• Online Access: https://hdl.handle.net/10356/146763
• Institution: Nanyang Technological University

Human and robot need to interact to perform complex tasks through mutual understanding. There are generally no effective means of communication from robot-to-human to express its ‘intention’. Effective recognition of a robot’s intention can mitigate potential conflicts or accidents. This is particularly important in the context of factories of the future where human and robot are expected to work in a shared environment. Beyond the relative structure of a factory, autonomous vehicles operate in shared environments with pedestrians and the ability of humans and autonomous vehicles to understand the other’s intention is crucial. In most environments, audio notifications, lack in effectiveness. The environment may be noisy and there is an additional problem of language and its understanding. Visual display may provide a better alternative. The use of traditional image projectors, for this purpose, has severe limitations in bright surroundings. Image brightness is crucial, and traditional video projection are typically confined to indoor usage or in the evening when the surrounding is relatively dark. In contrast, laser projection is seldom short of brightness, but there is a limit to the speed of vector drawing and power of commercially available laser projectors allowed for use in public space. Figure 1. Shape and text to be written by laser – A scenario for writing warnings on the floor from the 3m height mount frame. The use of “lower power laser” projectors are also disadvantaged by the size of the image projected and the complexity of the image as, (i) due to the scan rate of the galvanometer mirrors, it is difficult for a single laser writer to draw both graphics and text without flickering problem and (ii) a limitation of the angle of laser beam deflection that typically does not generally exceed ±20°. One solution is to use two lasers and split their tasks but negates their simplicity and cost factors. Such a graphic writer is also capable to display intentions, suggestions and warnings on any surface around, and thus inform humans of a robot’s actions and intentions. In this paper, a novel solution using a custom designed spatial Laser Graphic Projection (LGP) is presented for visual communication between robot and human. One inherent problem manifests itself with regards to the (relatively) low update rates due to the mechanical inertia of laser galvanometer mirrors used to direct the laser beam. The mirrors are specially designed to allow for dynamic manipulation of the laser beam. The inertia of the mirrors typically will limit the scan rate of the projected images. The design needs to respect the rotational resonance frequency of the mirror system to align with the typical scan rate needed to project images.