Title: Planning and Control for Multi-robot Manufacturing Processes
Date: Friday, August 15
Time: 12:00 pm – 2:00 pm
Location (in-person): GTMI 114
Location (remote): Microsoft Teams Link
Alex Arbogast
Robotics PhD Student
George W. Woodruff School of Mechanical Engineering
Georgia Institute of Technology
Committee:
Dr. Christopher Saldaña (advisor) - George W. Woodruff School of Mechanical Engineering
Dr. Thomas Kurfess - George W. Woodruff School of Mechanical Engineering
Dr. Jun Ueda - George W. Woodruff School of Mechanical Engineering
Dr. Joshua Vaughan - Manufacturing Science Division, Oak Ridge National Laboratory
Dr. Sean Wilson - Robotics and Autonomous Systems Division, Georgia Tech Research Institute
Abstract:
Multi-robot systems offer the potential to reduce lead times, improve fault tolerance, and broaden the capabilities of advanced manufacturing processes. The unprecedented complexities of modern component designs demand systems that synergistically integrate dexterous manipulation, precise coordination, and versatile process capabilities. Achieving this integration in multi-robot manufacturing systems requires sophisticated planning and control strategies, particularly in close-proximity environments. Despite growing industrial interest in convergent manufacturing processes, few studies have addressed the key challenges of tightly-coordinated motion control and heterogeneous process planning strategies for multi-robot manufacturing systems. Moreover, the impact of coordination strategies on process makespan and part quality remains poorly understood. This dissertation models a further understanding between coordination techniques and their impact on multi-robot production processes. Specifically, this work: (1) derives and analyzes control laws for coordinating kinematically redundant multi-robot systems within a shared task frame, (2) proposes novel methods for heterogeneous task allocation and scheduling under complex dependency and proximity constraints, and (3) establishes design principles and modeling guidelines to support the scalability of the heterogeneous task scheduling framework. The proposed methods are empirically validated on simulated and physical multi-robot systems, demonstrating reduced production times and practical geometric accuracies. This dissertation advances the theory and application of process planning and control in collaborative robotic manufacturing and provides a comprehensive investigation on the impacts within modern production processes.