Zhuolun (Alan) Zhao 赵卓伦

Robotics Master Student @ Penn | Robotics & AI Engineer

University of Pennsylvania GRASP (General Robotics, Automation, Sensing and Perception) Laboratory

Hi, I’m Alan!

I’m an aspiring engineer and entrepreneur building robots for the next generations. I believe robots, alongside humans, can make the world a better place.

I am interested in vision-based data-driven robot learning in human-robot hybrid environments. Currently a Master of Science in Robotics student at University of Pennsylvania, I am primarily working on robot learning from human research and open-source autonomous driving.

Previously, I obtained my BA/BS from Swarthmore College double majoring in Computer Science and Engineering. I have worked in the AI, robotics, and venture capital industries.

Interests

Artificial Intelligence
Robotics
Computer Vision
Machine Learning
Entrepreneurship

Education

M.S.E. in Robotics, 2024
University of Pennsylvania, Philadelphia, PA, USA
B.A. in Computer Science / B.S. in Engineering, 2018
Swarthmore College, Swarthmore, PA, USA
High School Diploma, 2014
The High School Affiliated to Renmin University of China, Beijing, China

Experience

Robotics Software Engineer

Autoware Foundation

October 2023 – Present Philadelphia, Pennsylvania

Building an autonomous gokart and deploy open-source autonomous algorithms
Won 1st Place at Autonomous Karting Series 2024

Research Analyst

Beijing Science & Technology Innovation Fund

August 2021 – June 2022 Beijing, China

Robotics and AI Industry Research

Robotics Software Engineer

Fetch Robotics Inc. (Acquired by Zebra Technologies)

March 2019 – June 2021 San Jose, California

I worked on:

Lane congestion detection
Human detection
Social navigation

https://fetchrobotics.com/fetch-robotics-blog/using-social-navigation-behaviors-to-optimize-amr-performance/

Projects

Open Source Autonomous Driving GoKart

We built and raced a self-driving GoKart.

3verFresh: Easy-to-use Herb Preserver

Keep herbs vibrant and kitchen chic

Footstep Planning and Control for Bipedal Robots using Reinforcement Learning

Teach a humanoid robot how to climb stairs

F1TENTH Autonomous Racing

Develop perception, machine learning, planning, control, mapping, and racing algorithm on a 1/10 autonomous driving platform.

Autonomous Block Stacking with Franka Emika Panda Arm Robot

Using computer vision and motion planning, grasp static and dynamic blocks and stack them on top with precision.

Investigating Neural Field Models For 3D-Aware GANs

Can a Gaussian-activated MLP replace the SIREN in pi-GAN?

Modeling Multimodal Distribution with Improved BicycleGAN

Improve BicycleGAN’s performance through architectures and loss function modifications

Comparing Evolutionary Algorithms for Deep Neural Networks

Using multi-layered coevolutionary approaches / hierarchical approaches to evolve deep neural network architecture (ICLR 2018 Reproducibility Challenge)

Featured Publications

Zhuolun (Alan) Zhao, Harsha Uppili, Gabriel Meyer-Lee

December, 2017 ICLR 2018 Reproducibility Challenge

Comparing Evolutionary Algorithms for Deep Neural Networks

The results of this paper show that both multi-layered coevolutionary approaches and hierarchical approaches to optimizing DNNs are feasible and can develop DNNs comparable to those with hand-designed architectures.