Dự án
Khám phá các dự án nghiên cứu và phát triển về UAV và Robotics
Sea Reconnaissance VTOL UAV – A Pioneering AI Solution for Vietnam’s Tourism Safety
URLAB introduces the Sea Reconnaissance UAV, a project developed by students at Ton Duc Thang University. With over 3,000 km of coastline and millions of visitors annually, Vietnam faces drowning risks from hazardous phenomena like rip currents and whirlpools. Traditional observation methods are limited in covering large areas and providing timely information. The Sea Reconnaissance UAV offers an advanced, specialized technological solution to address this challenge: • Integrates cutting-edge Artificial Intelligence (AI) (built on TensorFlow/PyTorch), enabling automatic detection of rip currents and whirlpools with 80-85% accuracy. The AI system is trained on real-world data from Vietnam’s coastal regions. • Capable of autonomous patrolling along pre-set routes, continuously monitoring vast sea areas, and delivering real-time data to digital maps (GIS) and mobile applications for tourists and rescue teams. • Equipped with 4K cameras and thermal imaging for detecting distressed individuals, with a flight endurance of 1 to 1.5 hours. • Designed with a corrosion-resistant carbon fiber frame, ensuring durability in harsh marine environments and all weather conditions. The UAV supports flexible takeoff and landing on beaches or small vessels.
GeoScout – High‑Performance 3D LiDAR Integration for Indoor UAV Logistics
Modern warehouse and logistics operations require accurate 3D perception in GPS‑denied indoor spaces, yet UAVs struggle with both high‑volume sensor data and environment constraints. GeoScout addresses these challenges by: • Integrating the SICK multiScan100/136 3D LiDAR with an NVIDIA Jetson Orin 8 GB embedded computer under ROS 2 Humble. • Providing a custom C++ driver that receives raw UDP packets, deserializes proprietary MSGPACK streams (via msgpack‑c), transforms polar coordinates (range, azimuth, elevation) into Cartesian (X, Y, Z), and publishes standard sensor_msgs/msg/PointCloud2 messages on `/lidar/points`. • Achieving low resource usage on the Jetson Orin (total system load ∼20% CPU, ∼165 MB RAM) while delivering real‑time point‑cloud visualization in RViz2.
.jpg&w=3840&q=75)
Research and Development of an Autonomous Drone System Integrating Computer Vision for Target Recognition and Precision Payload Deployment
This project focuses on the research and implementation of an autonomous drone system capable of executing fully automated flight missions in GPS‑denied environments. The drone is equipped with an IMU, altimeter, optical flow module, and camera to enable relative localization and maintain stable flight trajectories. The system incorporates real‑time image processing to detect, localize, and track specific targets. Once a target’s relative position is identified, a precision payload release mechanism is triggered via a tailored dynamic model and high‑precision control loop. From navigation through target recognition to payload deployment, all processes are managed by a dedicated autonomous control unit designed for strict safety and accuracy requirements. Applications include search and rescue operations, indoor material transport, and competition entries in intelligent autonomous systems.
.jpg&w=3840&q=75)
Design and Development of a Fixed-Wing Tricopter Drone System for Agile Navigation and High Flight Efficiency
This project presents a hybrid drone platform combining a fixed‑wing airframe with a three‑rotor (tricopter) propulsion system to achieve both long‑range efficiency and VTOL agility. Key design and development features: • Aerodynamic Optimization: Lightweight yet robust composite frame, streamlined wing profile, and variable‑pitch tail rotor for stability in hover and precision in forward flight. • Hybrid Propulsion: Three‑rotor tricopter configuration for vertical takeoff/landing and agile yaw control, seamlessly transitioning to fixed‑wing cruise for high‑efficiency forward flight. • Advanced Control: Nonlinear flight‑control algorithms fusing IMU, GPS, and altimeter data to maintain stable trajectories across wind disturbances and complex terrain. • Versatile Mission Profiles: Capable of terrain mapping, lightweight cargo delivery, wide‑area surveillance, and dynamic hover‑to‑cruise mode switching for rapid response tasks.
MINI DRONE STEM – Development of Mini Drone STEM Models for Education
This project is dedicated to the research, development, and application of mini drone STEM (Science, Technology, Engineering, and Mathematics) models as a groundbreaking educational tool. Our core objective is to harness the inherent appeal and interactive nature of drones to ignite a passion for learning, while strengthening students’ logical thinking, problem‑solving skills, and creativity. Key elements: • Compact, affordable, and highly customizable drone designs that students at various skill levels can easily assemble, program, and pilot. • Development of comprehensive curricula and teaching materials covering aerodynamic principles, mechanical design, electronics, basic programming, and real‑world drone applications. • Interactive lessons and guided exercises that reinforce core STEM topics through drone‑based experiments and projects. • A scalable platform that supports progression from beginner kits to advanced modules, empowering schools and learning centers to implement drone‑based STEM activities effectively.
STEM SPIDER – Educational Robot with Vietnamese Culture
The four‑legged, spider‑inspired robot features a stylized conical hat and golden star pattern on red—icons of Vietnamese heritage. STEM SPIDER serves both as a hands‑on STEM education tool and as a celebration of creativity and national pride. Control System: • Central Controller – ESP32: Powerful microcontroller with built‑in Wi‑Fi and Bluetooth; offers flexible, extensible programming ideal for education and research. • Ultrasonic Sensor: Detects and measures distance to obstacles; enables collision avoidance for safe, stable movement in real‑world environments. • AI Thinker: Embeds lightweight AI models on the robot for voice recognition and basic command responses; interacts with students via sound, light, and motion signals to foster an engaging learning experience.
.jpg&w=3840&q=75)
Research and Development of an Interactive Ant-Inspired Robot Model Using 3D Printing Technology and ESP32 Microcontroller
This project designs and fabricates an interactive, ant-inspired robot using 3D‑printed components for lightweight flexibility and customizability. Powered by an ESP32 microcontroller, the system controls multi‑degree servo motors at each leg joint to mimic real ant locomotion patterns. Embedded motion‑control algorithms enable the robotic ant to walk, turn its head, and respond to environmental stimuli via onboard sensors. Developed for insect locomotion research and STEM education, the platform also serves as a testbed for bio‑inspired robotic systems. Future enhancements include inter-robot communication, additional environmental sensors, and collective behavior simulation (foraging, pheromone routing, cooperative group dynamics).
Development of a Real-Time Thermal Video Streaming Drone System over 5G Internet Connectivity
This project develops a drone system that enables real-time thermal video streaming over a 5G network, optimized for remote monitoring in complex or hazardous environments. Core components: • Thermal Gimbal Camera: Captures stable infrared imagery for detecting heat signatures in real time. • Raspberry Pi 5: Acts as the onboard processing and streaming unit, handling video encoding and transmission. • 5G SIM Module: Provides high-speed, low-latency internet connectivity without relying on local Wi-Fi infrastructure. • Web-Based Interface: Allows users to remotely view live thermal feeds, store footage, and engage in two-way communication with the system. Target applications include wildfire detection, search and rescue operations, industrial inspections, and security/defense missions.
Design of an Autonomous Quadruped Robot for Complex Terrain Navigation
This project focuses on designing and fabricating a quadruped robot capable of agile and stable locomotion across complex terrains. With high-precision servo motors and an integrated Inertial Measurement Unit (IMU), the robot maintains balance and adapts its posture in real-time based on environmental feedback. Key features include: • Gait Algorithms: Implements various gait patterns such as trot, crawl, and wave gait to adapt to different surfaces and movement strategies. • Terrain Adaptability: Capable of walking forward/backward, turning, and overcoming small obstacles, while maintaining stability against external forces. • Research Integration: Serves as a platform for biomechanics research, autonomous navigation, and real-time control experiments. 🔧 Structure and Operating Principles **1. Mechanical Components** • **Main Body Frame System**: Central chassis that houses control boards and connects to all limbs. • **Leg Locomotion Assembly**: Mechanical structure comprising joints, servo actuators, and linkages for each leg. • **Central Controller**: ESP32-WROOM-32 microcontroller providing wireless connectivity and real-time control. **2. Software and Electronics** • **Development Tools**: Arduino IDE and VS Code used for firmware development. • **Sensors**: Includes a 10-DOF IMU (ADXL345, ITG3200, VCM5883L, BMP085) for attitude estimation. Future plans include LiDAR or 3D vision for localization and obstacle perception. • **Programming Framework**: Utilizes ROS (Robot Operating System) for motion control, sensor integration, and algorithm development. **3. Algorithms** • **Kinematics**: Robot links are defined using Denavit-Hartenberg (D-H) parameters to solve forward and inverse kinematics. • **Transformation Matrices**: Homogeneous transformation matrices are applied to determine the position and orientation of the robot's limbs in space. • **Control Strategy**: PID or trajectory-based controllers ensure smooth joint motion and posture stabilization. 📈 Applications • STEM education and robotics courses. • Biomechanics and multi-legged locomotion research. • Development of AI-based autonomous robotics platforms. 📚 References • DFRobot 10-DOF IMU Wiki (V2.0 & V1.0) • Sensor datasheets: ADXL345, ITG3200, VCM5883L, BMP085, BMP280 • ROS Documentation & Arduino-ROS libraries
Lĩnh vực nghiên cứu
Các lĩnh vực nghiên cứu chính của phòng URLab
Thiết kế và chế tạo UAV
Nghiên cứu và phát triển thiết kế UAV mới với hiệu năng và khả năng cải tiến.
Hệ thống điều khiển tự động
Phát triển thuật toán điều khiển cho UAV tự động và bán tự động.
Xử lý ảnh và cảm biến
Nghiên cứu kỹ thuật xử lý ảnh và tích hợp cảm biến cho UAV.
Ứng dụng cứu hộ
Phát triển giải pháp giám sát ven biển để phát hiện xoáy nước, dòng chảy xiết và hỗ trợ tìm kiếm cứu nạn trong rừng.
Giám sát môi trường
Ứng dụng UAV để giám sát và bảo vệ môi trường, bao gồm theo dõi ô nhiễm và bảo tồn đa dạng sinh học.
Giáo dục STEM
Phát triển bộ kit Drone và Robotics cho học sinh, sinh viên và người học trẻ.