Enabling automated facility maintenance from articulated robot Collision-Free designs. (January 2023)

• Record Type: Journal Article
• Authors: Vann, William; Zhou, Tianyu; Zhu, Qi; Du, Eric
• Is Part Of: Advanced engineering informatics. Volume 55(2023)
• Publication Date: 2023-01
• Subjects: Facility maintenance -- Human-Robot Collaboration -- Collision -- Monte Carlo Simulation; Computer-aided engineering -- Periodicals; Engineering -- Data processing -- Periodicals; 620.00285
• DOI: 10.1016/j.aei.2022.101820
• Languages: English
• ISSNs: 1474-0346

Abstract

Abstract: Key challenges found in some facilities maintenance (FM) operations include the dependency on placing sophisticated workers in undesirable working conditions that could make them susceptible to accidents. Robot-assisted in-situ facilities maintenance represents new opportunities by assisting human operators when exposed to harsh environments. However, given the complexities and potential varieties of advanced robotic machinery available for use in facility maintenance workplaces in the future, there is still a legitimate risk of collisions between future robotic systems and the facility structure during the robot-assisted processes. This paper proposes and tests a simulation-based collision-free design method for future facilities that may be benefited from a robot-assisted FM process. At the most fundamental level, we consider interactions between the articulated manipulator (robot) and the surrounding facility structure. Inverse Kinematics (IK) and game engines are used to simulate the possible collisions between a given robot and the parameterized facility layout. To account for the uncertainty about the kinematic specifications of the future robotic machinery, a Monte Carlo method is used to model unique priority circumstances for each joint on the articulated manipulator. The simulation result, presented as the robot work zone envelope, is then used to estimate the collision probability given a certain design parameter, and the corresponding optimal design that balances the initial construction cost and the FM costs. The method is assessed for identifying a 3D spatial collision probability cloud within reach of a 7 DOF articulated manipulator that is marginally positioned within a pressure valve piping facility. These insights can provide future facility designers, managers, and equipment operators with a quick and flexible 3D spatial collision probability indicator to better design the facility and proximity limitations of any articulated manipulator positioned throughout a confined space for future robot-assisted FM processes.