My Robotics Journey
From my first encounter with a VEX robot in high school to conducting cutting-edge research at UCLA's Robot Intelligence Lab, my journey in robotics has been a story of passion, perseverance, discovery, and, most importantly, love.
The Spark: LEGO NXT, EV3, and First LEGO League (FLL)
At the age of 8, I attended a talent show where I was performing martial arts. The boy next to me was showcasing LEGO NXT robots he had assembled, and I was instantly captivated. That moment sparked a fascination that would shape my future.
I begged my parents to let me learn robotics, and soon after, I joined the Nanjing Youth Palace summer program in 2012. This marked the beginning of my journey into robotics structures and programming. Over the next three years, I participated in multiple seasons of the First LEGO League (FLL), where I learned the fundamentals of design, programming, and teamwork through hands-on robotic challenges.
National Awards in World Educational Robots Competition (WER)
My passion for robotics competitions grew, and I began competing nationally in the World Educational Robots Competition (WER). This experience taught me the discipline of iterative design and precise programming.
I dedicated countless hours to testing, assembling, coding, and debugging line-following robots that could intelligently complete complex tasks. My hard work paid off with a National First Prize in 2015 and a Silver Medal (Finalist) in 2017, achievements that fueled my growing passion for competitive robotics.
High School VEX Robotics Club
As President of the VEX Robotics Club at Jinling High School, I led a team of 15 students in designing, building, and programming robots for competitive challenges. This experience taught me the fundamentals of mechanical design, sensor integration, and autonomous programming.
I participated in three VEX seasons: Change Up (2020-2021), Tipping Point (2021-2022), and Spin Up (2022-2023). Over these seasons, I built approximately 10 robots, with three primary designs each year. In Spin Up, two of my team's innovative designs can be viewed here and here.
Our team achieved remarkable success, winning first prize in the VEX World Championship 2021, first prize in the VEX Asia Championship 2023, and reaching the national finals in multiple VEX Robotics Competitions. We implemented innovative solutions for object manipulation and autonomous navigation tasks. These early successes fueled my passion for robotics and solidified my decision to pursue engineering at the university level.
Key Learnings:
- Team leadership and project management in technical settings
- Fundamentals of robot kinematics and control systems
- Problem-solving under competition pressure
High School Research and University Foundations
During my final two years of high school, I embarked on a self-directed journey into neural networks and artificial intelligence in computer vision. I researched innovative combination methods for semantic segmentation specifically for surgical robots, exploring how AI could enhance precision in medical procedures.
Entering UCLA's Computer Engineering program opened new horizons. Courses like Introduction to Robotics (CS 188) and Computational Imaging (ECE 239AS) provided the theoretical foundation I needed to understand complex robotic systems at a deeper level, bridging the gap between my hands-on experience and academic theory.
Research Breakthrough: Long-Horizon Imitation Learning
My current research at the UCLA Robot Intelligence Lab (URIL) under Professor Yuchen Cui focuses on long-horizon robotics imitation learning. We're tackling one of the most challenging problems in robotics: enabling robots to learn complex, multi-step tasks from limited demonstrations.
Our innovative approach combines complex inputs (such as gaze tracking) with transformer-based architectures to enable robots to understand and execute tasks that span hundreds of actions over extended time horizons.
Research Focus Areas:
- Multi-modal learning: Combining vision, gaze, language, and tactile inputs
- Temporal abstraction: Learning high-level skills that can be composed
- Interpretability: Understanding and interpreting AI controller outputs
- Sample efficiency: Reducing the number of demonstrations needed
- Generalization: Applying learned skills to novel environments
What excites me most about this research is its potential for real-world impact. From assistive robotics that can help with daily tasks to industrial automation that adapts to changing environments, the applications are vast and meaningful. We're not just teaching robots to perform tasks—we're teaching them to understand and adapt.
Looking Ahead: The Next Frontier
As I look toward the future, I see several exciting directions for my work in robotics. Each represents a frontier where technology meets human need, and where intelligent systems can enhance our capabilities and improve our world.
Embodied AI
Developing robots that can learn and adapt in unstructured environments through embodied experience, creating machines that understand the world not just through data, but through interaction.
Neuro-Symbolic Systems
Combining neural networks with symbolic reasoning for more interpretable and reliable robotic decision-making, bridging the gap between learning-based approaches and rule-based systems.
Robotics for Good
Applying robotics to address societal challenges in healthcare, environmental monitoring, and accessibility, ensuring that technological advancement serves humanity's greatest needs.
My journey in robotics—from childhood fascination to cutting-edge research—has taught me that technology is most powerful when driven by purpose and guided by human values. Each challenge overcome and each discovery made reinforces my belief that intelligent robots, thoughtfully developed, will play a crucial role in shaping a better future for humanity.