Hi There ~

I am a Software Development Engineer at Amazon Web Services (AWS). I received my Master's degree from the University of Southern California (USC), majoring in Computer Science. I received my B.Eng. in Electrical Engineering with honor from Chongqing University (CQU), China. From March 2022 to December 2022 and January 2024 to May 2024, I worked as a robotics software developer in the RROS Lab at the USC Center for Advanced Manufacturing (CAM). From January 2023 to May 2023, I worked as a Software Development Engineer Intern at Futronics. From May 2023 to August 2023, I worked as an SDE intern at AWS.

During my undergraduate study, my research mainly focused on power systems, circuit design, and embedded system programming. For instance, the appliance of Deep Reinforcement Learning (DRL) and Multi-Agent theory in a wind farm.

Mail: peijiexu99@gmail.com | peijiexu@usc.edu

Links

Projects   |   Linkedin   |   CV (2022) (pdf)   |   Articles

Recent Projects

• Collected functional requirements from lab researchers, formulated the problems, analyzed and implemented solutions
• Refined 2 Robot Operating System (ROS) infrastructures for controlling manipulators (KUKA iiwa & ABB) to meet the lab's needs, designed and implemented C++ interfaces for the KUKA infrastructure with detailed documentation
• Developed 5 required functionalities (multi-robot collaboration with a distributed network, jerky motion issue elimination, frame management system for manipulators, software-level emergency stop service, impedance/position control mode switching service) and 2 extra features (rapid ROS Java message generation, robot state visualization/monitor)
• Leveraged Docker for rapid ROS Java message generation, reducing the generation time from 17 to 2 minutes
• Devised a robot state visualization/monitor service to show the real-time dynamics of the robot, and introduced the watchdog daemon (a multi-thread-based protection mechanism) to handle the case that the corresponding client crashes unexpectedly
• Improved the pipeline and algorithm for computing the velocity of the end-effector, reduced inter-process communication loads, and decreased the response time from 130 ms to less than 1 ms

• Applied Multi-Agent DRL to minimize the Average Voltage Deviation of the offshore wind farm
• The work was awarded the outstanding undergraduate thesis of CQU (top 1.5 % in the university)
• Detail…

• Controled a humanoid robot to finish a sequence of challenging tasks
• Detail…

• Built a C++ library to realize programmable control of an old Denso manipulator, then migrated the library to ROS
• Detail…

Next: Project List

Back to Top