Digital Twin of Industrial Robots Over the Cloud: Framework, Design and Analysis

Yuktha M,Ashok Kumar Patil,Aakash Negi,Hari Ganesan,Aaryan M Sandeep

DOI: https://doi.org/10.1109/ICRAS62427.2024.10654490

2024-06-21

Abstract:This paper presents a novel approach for creating real-time, secure and interactable Digital Twins of industrial robots. This approach diverges from previous approaches that relied on Digital Twin Definition Language (DTDL). The proposed method uses Robotics Operating Systems (ROS) and the Unity Game Engine to build the Digital Twins. The prototype built to showcase this approach comprises 4 sub-systems namely Data Acquisition sub-system, Cloud sub-system, Digital Client sub-system, and Remote-Control sub-system. The methodology focuses on real-time data synchronization using cloud-based communication services like loT servers. There are also multiple security measures employed to secure the data- like passing data through proxy and end-to-end encryption of data. The prototype made using Niryo Ned 2 Robot and a Raspberry Pi 3b+ showcases accurate physical-Digital Twin and effective remote control. Key findings highlight the system's ability to overcome latency, security and data accuracy, which in turn helps improve human-robot collaboration and interactions. This research contributes to the Digital Twin field by providing a modular and scalable framework that can be adapted for various robots and systems. This paper not only provides the general idea for implementation and problems faced during development but also suggests pathways for future work, like - optimization of message rates, integration of machine learning for maintenance and other purposes.

Computer Science,Engineering