Date of Award

May 2021

Document Type


Degree Name

Doctor of Philosophy (PhD)


Automotive Engineering

Committee Member

Venkat Krovi

Committee Member

Bing Li

Committee Member

Matthias Schmid


In the past decades, automation in the automobile production line has significantly increased the efficiency and quality of automotive manufacturing. However, in the automotive assembly stage, most tasks are still accomplished manually by human workers because of the complexity and flexibility of the tasks and the high dynamic unconstructed workspace. This dissertation is proposed to improve the level of automation in automotive assembly by human-robot collaboration (HRC). The challenges that eluded the automation in automotive assembly including lack of suitable collaborative robotic systems for the HRC, especially the compact-size high-payload mobile manipulators; teaching and learning frameworks to enable robots to learn the assembly tasks, and how to assist humans to accomplish assembly tasks from human demonstration; task-driving high-level robot motion planning framework to make the trained robot intelligently and adaptively assist human in automotive assembly tasks. The technical research toward this goal has resulted in several peer-reviewed publications. Achievements include: 1) A novel collaborative lift-assist robot for automotive assembly; 2) Approaches of vision-based robot learning of placing tasks from human demonstrations in assembly; 3) Robot learning of assembly tasks and assistance from human demonstrations using Convolutional Neural Network (CNN); 4) Robot learning of assembly tasks and assistance from human demonstrations using Task Constraint-Guided Inverse Reinforcement Learning (TC-IRL); 5) Robot learning of assembly tasks from non-expert demonstrations via Functional Objective-Oriented Network (FOON); 6) Multi-model sampling-based motion planning for trajectory optimization with execution consistency in manufacturing contexts. The research demonstrates the feasibility of a parallel mobile manipulator, which introduces novel conceptions to industrial mobile manipulators for smart manufacturing. By exploring the Robot Learning from Demonstration (RLfD) with both AI-based and model-based approaches, the research also improves robots’ learning capabilities on collaborative assembly tasks for both expert and non-expert users. The research on robot motion planning and control in the dissertation facilitates the safety and human trust in industrial robots in HRC.



To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.