🔬 Active NSTC Projects
Development of a High-Performance Autonomous Navigation System for Unmanned Vehicles Integrating Fiber Optic Gyroscope Inertial Navigation and Multi-Modal AI Visual Perception
Project Overview:
Inspired by human autonomous motion and decision-making processes, this project aims to develop an AI-based navigation system for unmanned vehicles operating across maritime, ground, and aerial domains. By emulating the spatial perception and motion-planning capabilities of human operators, the proposed system is designed to achieve deeper environmental understanding and more intelligent autonomous decision-making. The project is expected to contribute substantive innovations to autonomous vehicle research and support the advancement of Taiwan's all-domain unmanned vehicle industry.
Inspired by human autonomous motion and decision-making processes, this project aims to develop an AI-based navigation system for unmanned vehicles operating across maritime, ground, and aerial domains. By emulating the spatial perception and motion-planning capabilities of human operators, the proposed system is designed to achieve deeper environmental understanding and more intelligent autonomous decision-making. The project is expected to contribute substantive innovations to autonomous vehicle research and support the advancement of Taiwan's all-domain unmanned vehicle industry.
Digital Twin Cross-Platform Collaborative R&D for Intelligent Automation of Container Terminal Vehicles
Project Overview:
National Taiwan University and Chenxi Precision have jointly established the Building and Port Automated Vehicle R&D Sub-Center, which focuses on the automation and autonomy of five categories of industrial vehicles. Built upon a digital twin framework, the project encompasses drive-by-wire retrofitting, remote operation, human–machine control switching, multimodal perception, and cross-platform fleet coordination and dispatch. The project also includes international research collaboration with the University of British Columbia (UBC), Canada.
National Taiwan University and Chenxi Precision have jointly established the Building and Port Automated Vehicle R&D Sub-Center, which focuses on the automation and autonomy of five categories of industrial vehicles. Built upon a digital twin framework, the project encompasses drive-by-wire retrofitting, remote operation, human–machine control switching, multimodal perception, and cross-platform fleet coordination and dispatch. The project also includes international research collaboration with the University of British Columbia (UBC), Canada.
SDN-Driven UAV Swarm Control: SIL/HIL Simulation Platform Architecture and Networked Control Performance Validation
Project Overview:
This project develops an intelligent command and control framework with advanced environmental understanding and decision-making capabilities, enabling a single operator to intuitively coordinate complex UAV swarms. Through the SAVLink semantic-layer communication protocol, natural-language commands are translated into precise logical instructions, improving bandwidth utilization and enhancing system resilience in interference-prone environments. The project also integrates Software-in-the-Loop (SIL) and Hardware-in-the-Loop (HIL) simulation to establish and validate a platform for UAV swarm missions, networked control, and communication performance evaluation.
This project develops an intelligent command and control framework with advanced environmental understanding and decision-making capabilities, enabling a single operator to intuitively coordinate complex UAV swarms. Through the SAVLink semantic-layer communication protocol, natural-language commands are translated into precise logical instructions, improving bandwidth utilization and enhancing system resilience in interference-prone environments. The project also integrates Software-in-the-Loop (SIL) and Hardware-in-the-Loop (HIL) simulation to establish and validate a platform for UAV swarm missions, networked control, and communication performance evaluation.
📋 Other Active Projects
Hardware-in-the-Loop (HIL) Research for UAV Hybrid Propulsion Systems
Project Overview:
This project aims to enhance the design, modeling, and testing capabilities of hybrid propulsion systems for unmanned aerial vehicles. Its primary objectives include establishing the overall system architecture and mathematical models, as well as developing a Hardware-in-the-Loop (HIL) testing workflow for validating energy-management and control strategies. The research methodology encompasses system-level analysis, HIL platform development, and controller interface integration.
This project aims to enhance the design, modeling, and testing capabilities of hybrid propulsion systems for unmanned aerial vehicles. Its primary objectives include establishing the overall system architecture and mathematical models, as well as developing a Hardware-in-the-Loop (HIL) testing workflow for validating energy-management and control strategies. The research methodology encompasses system-level analysis, HIL platform development, and controller interface integration.
🌍 International Collaboration
Physics-Informed Neural Network-Based Material Cyclic Hardening Prediction and Assessment
Project Overview:
In collaboration with Tohoku University, Japan, this project develops a prediction framework based on Physics-Informed Neural Networks (PINNs). By extracting information from monotonic tensile test data, the framework inversely identifies complex cyclic hardening parameters while incorporating governing equations from solid mechanics to ensure physical consistency. Monte Carlo simulation is further integrated with finite element analysis to quantify how material-level uncertainties affect macroscopic structural behavior.
In collaboration with Tohoku University, Japan, this project develops a prediction framework based on Physics-Informed Neural Networks (PINNs). By extracting information from monotonic tensile test data, the framework inversely identifies complex cyclic hardening parameters while incorporating governing equations from solid mechanics to ensure physical consistency. Monte Carlo simulation is further integrated with finite element analysis to quantify how material-level uncertainties affect macroscopic structural behavior.