Iβm a first-year M.S. student from Department of Mechanical Engineering, Columbia University. My research interest includes robotics, robot manipulator, artificial intelligence, control and cloud-based vehicle control.
You can find my CV here: Taoβs Curriculum Vitae.
LinkedIn: linkedin.com/in/naian-tao
π Educations
Mechatronics Modeling & Simulation
- Aug 2024 β Present. Columbia University, New York, NY, US
- M.S. in Mechanical Engineering (Robotics Track)
- Courses: Robotics Studio, Reinforcement Learning, Control
- Aug 2023 - Jun 2024. University of Detroit Mercy, Detroit, MI, US
- B.E. in Mechatronics, Robotics, and Automation Engineering, GPA: 3.81/4.00
- Courses: Robotics, Autonomous Mobility Robotics, Embedded Systems, Computational Intelligence Technique
- Sept 2020 - Jun 2024. Beijing University of Chemical Technology, Beijing, CN
- B.E. in Mechanical Design, Manufacturing and Automation, GPA: 3.62/4.33
- Courses: Automatic Control Design, Artificial Intelligence, Mechine Design, Program Design, Hydraulic and Atmospheric Pressure Transmission
π Projects
Robotics Senior Design at the University of Detroit Mercy
- Designed an assistant robot system to help elderly individuals, combining mobile robots, computer vision, and a robotic manipulator.
- Modeled the ReactorX-200 robotic arm and developed an app in Matlab to simulate both forward and inverse kinematics of the robotic arm. Finally, applied the algorithm to the real robotic arm for verification and used the program to control the arm.
- Trained dataset using YOLOv8 and combined it with Kinect V2 point cloud data to achieve target object localization.
- Performed camera calibration and hand-eye calibration between the robotic arm and the camera, significantly improving the grasping accuracy.
- Implemented navigation functionality using the A* algorithm for global path planning and the DWA algorithm for local path planning.
RoboCup China Open 2022 ROBOCUP@HOME
-
Designed home service robot with capabilities in guest reception and guidance, object grasping & delivery, object recognition, and voice interaction.
-
Designed and executed a algorithm combining object recognition with object grasping and delivery.
-
Orchestrated a four-step approach, including precise identification of target objects, accurate localization using depth cameras, pose analysis for gripping, and precise control of the robotic arm for successful object grasping and delivery.
-
Utilized MoveIt with Open Motion Planning Library (OMPL) for motion and path planning, RRT* algorithm for efficient trajectory generation, and TRAC-IK for accurate inverse kinematics.
-
Integrated YOLOv5 and Kinect DK to achieve target object recognition and 3D localization.
-
Used the Grasp Pose Detection (GPD) package to detect 6-DOF grasp poses for a 2-finger robot hand in 3D point clouds, enabling the grasping of objects in various orientations.
-
Successfully developed and implemented a unique robotic door-opening solution to complete the challenging task β the only team that completed this task.
-
Innovatively combined precise base positioning and mechanical arm path planning, significantly reducing computation time, streamlining the process, and efficiently completing door-opening tasks while ensuring obstacle avoidance.
RL Projects
- Complete the CartPole task in gym using the Actor-Critic Policy Gradient algorithm and the Deep Q-Network (DQN) algorithm, respectively.
- Successfully applied the Q-Learning algorithm to solve the Frozen Lake environment.
π» Internships
- Feb 2023 - May 2023, Tsinghua University Intelligent Connected Vehicle Research Group, Beijing, China.
- Developed a system integrating electric truck fleet scheduling and battery swapping optimization with cloud control, enhancing efficiency and reducing energy waste. In simulation experiments, using this algorithm saves $15\%$ more energy and $9\%$ more cost compared to the Cruise Control (CC) method.
- Created a Predictive Cruise Control (PCC) for cruise control in electric trucks. Factored in road slope information and used the Dynamic Programming (DP) algorithm to generate efficient speed sequences, reduce energy consumption, and enhance energy regenerative capabilities.
- Implemented cloud control for real-time fleet, environment, and infrastructure connectivity, improving adaptive driving with dynamic traffic data by using the Genetic Algorithm (GA).
- Co-authored a patent on Battery Swapping Rhythm Planning and Predictive Cruise Control Method for Electric Heavy Truck Fleets (Patent Number: CN117002500A).
π Honors and Awards
- Jan 2023, Individual Scholarship in School Year 2021-2022 Term 2 in BUCT
- Nov 2022, National First Prize in 2022 RoboCup China Open ROBOCUP@HOME
- Nov 2022, Second Prize in 2021 China Robot Skills Competition, Beijing Division A Group
- Nov 2022, Qualified in the program Intelligent Four-legged Voice Interactive Robot in 2022 Innovation and Entrepreneurship Training Program for college students
- Nov 2022, Qualified in the program Humanoid Soft Robot in 2022 Innovation and Entrepreneurship Training Program for college students
- May 2022, Third Scholarship in School Year 2021-2022 Term 1 in BUCT
- Dec 2021, National Second Prize in 2021 China Robot Skills Competition
- Dec 2021, Third Scholarship in School Year 2020-2021 Term 2 in BUCT
- Jul 2021, Third Prize in 2021 BUCT Intelligent Service Robot Competition
- May 2021, Third Scholarship in School Year 2020-2021 Term 1 in BUCT
- Apr 2021, Outstanding Student Leader in BUCT CEE
π§ Skills
- Programming Language: Proficient in using C++, Python, and Matlab in robotics.
- ROS: Proficient in working with ROS.
- Robotics and Robot Control: Proficiency in using robotics control frameworks like MoveIt for motion planning. Skill in kinematics solving.
- Navigation and Mapping: Advanced skills in 2D indoor mapping using tools like Google Cartographer. Implementation of path planning algorithms and tools, including DWA, RRT, A*, Move_Base, etc.
- Cloud-Based Vehicle Control: Hands-on experience in developing cloud-based control systems for vehicles. Algorithm development for optimizing vehicle speed and power management.
- Microcontroller Proficiency: Proficient in the use of TM4C microcontrollers and Arduino.
- Mechanical Design: Proficient in manual engineering drawing and using SolidWorks for mechanical design, having completed multiple mechanical design projects.
- Mathematical Modeling: Deep understanding of mathematical modeling methods and tools. Application of mathematical modeling principles in real-world problem-solving.
- Soft skills: Quick learning ability, self-starting, and problem-solving as demonstrated in numerous competitions and internships.