Digital Twin Model Development for Autonomous Tram Localization Yolanda Tania Mulyadi (a*), Muhammad Rifqi Rafian Putra (a), Yul Yunazwin Nazaruddin (b), Miranti Indar Mandasari (c)
a) Department of Engineering Physics, Institut Teknologi Bandung
*yolanda.tania0803[at]students.itb.ac.id
b) Instrumentation and Control Research Group, Department of Engineering Physics, National Center for Sustainable Transportation Technology, Institut Teknologi Bandung
c) Building Physics Research Group, Department of Engineering Physics, Institut Teknologi Bandung
Abstract
The rapid advancement of information technologies led to rapid growth of various aspects, one of which is autonomous vehicles. Digital twin technology is one that is being frequently developed in autonomous vehicles research, enabling real-time remote monitoring and control of the vehicle^s physical asset. This technology not only reduces the maintenance costs and risks, but also prevents and speeds up accident management. This paper proposes a digital twin model for the autonomous tram, one of the vehicles that is being widely explored due to its safety, low emissions, and high capacity. In this research, the proposed digital twin model is utilized to virtually represent the kinematics of tram prototype in 2D model from data sent via Message Queuing Telemetry Transport (MQTT) protocol, enabling real-time remote control with low-band consumption. Virtual representation of the tram prototype is gathered via physical sensors and the virtual sensor consists of an Unscented Kalman Filter (UKF) and Long Short-Term Memory (LSTM) and controlled by Stanley controller. The results confirmed that the use of the proposed digital twin model could remotely monitoring and control the autonomous tram prototype in real-time condition.
Keywords: autonomous tram- digital twin- message queuing telemetry transport- virtual sensor- Stanley controller
