PhD defence by Junru Ren

Abstract

Digitalization includes a wide range of topics and applications. In this thesis, robots and advanced digital technologies are integrated to transform traditional processes into automated, data-driven systems. At the same time, advanced technologies such as sensor fusion, artificial intelligence (AI), and Digital Twin (DT) enable robots to bridge the physical and digital realms, creating smarter, more connected systems capable of self-monitoring and optimization.

From manufacturing to laboratories, these systems are increasingly relied upon to enhance productivity, safety, and efficiency. However, effective deployment of autonomous systems in real-world applications requires more than optimizing their individual performance. To ensure seamless integration, it demands a comprehensive understanding of their interactions with interconnected subsystems, their impact on surrounding environments, and their requirements. Monitoring these systems is critical to maintaining operational integrity, improving system safety, and enhancing autonomy.

Traditionally, such monitoring tasks have depended on human supervision, limiting scalability and efficiency. This thesis explores the solutions to automated monitoring systems through the implementation of vision-based algorithms. The research presented in this project addresses challenges in system function design and system monitoring. To address the challenge of system function design, a systematic framework is developed to identify functional requirements and potential hazards in autonomous systems while considering the interaction with surrounding objects. To tackle the challenge of system monitoring, different monitoring systems are designed to support risk assessment, remote monitoring, and operational status evaluation. These systems employ traditional computer vision algorithms alongside advanced deep-learning techniques, enabling real-time, precise, and flexible monitoring processes.

Applications explored in this thesis include safe human-robot collaboration, DT development, and status monitoring in laboratory automation. By addressing these challenges, this thesis contributes to the broad field of autonomous system digitalization. The proposed methodologies enhance the safety and functionality of individual systems and promote their integration into complex environments, supporting a future where robots and autonomous systems play a central role in operation digitalization.

Supervisors

• Main supervisor: Professor Lazaros Nalpantidis, DTU Electro, Technical University, Denmark
• Co-supervisor: Professor Ole Ravn, DTU Electro, Technical University, Denmark
• Co-supervisor: Professor Emeritus Nils Axel Andersen, DTU Electro, Technical University, Denmark

External examiners

• Associate Professor Fredrik Asplund, KTH – Royal Institute of Technology, Sweden
• Associate Professor Dimitrios Chrysostomou, Aalborg University, Denmark

Chair of assessment committee

• Associate Professor Yang Zhang, Technical University, Denmark

Master of the Ceremony

• Associate Professor Evangelos Boukas, Department of Electrical and Photonics Engineering, DTU