Guillermo Castillo

Guillermo Castillo

Ph.D. Canditate in Electrical and Computer Engineering

The Ohio State University - Cyberbotics Lab

I am a Ph.D. candidate at the Cyberbotics Lab at The Ohio State University with six years of hands-on experience as a Robotics Engineer. My expertise lies in creating robuts locomotion controllers for bipedal locomotion through the combination of reinforcement learning and model-based control in both simulations and real-world hardware scenarios.

Interests
  • Robotics
  • Control Systems
  • Reinforcement Learning
  • Machine Learning
Education
  • PhD in Electrical and Computer Engineering, current

    The Ohio State University

  • MSc in Electrical and Computer Engineering, 2019

    The Ohio State University

  • BSc in in Automation and Control Engineering, 2015

    Escuela Politecnica Nacional, Ecuador

Experience

 
 
 
 
 
The Ohio State University
Graduate Research Associate
August 2017 – February 2024 Columbus
  • Worked in system identification, model-based controller design, reinforcement learning, data-driven control, and hardware implementation of controllers for bipedal locomotion (sim-to-real transfer).
  • Assisted in tasks for lab management, successfully procuring new resources, creating and maintaining code repositories, designing hardware experiments, and coordinating team meetings and demos for lab visitors.
  • Mentored undergraduate and graduate students at the lab with research projects and honors theses.
 
 
 
 
 
Escuela Politecnica Nacional
Full-time lecturer and lab instructor
September 2015 – August 2017 Quito-Ecuador

Instructed undergraduate courses to groups of up to 50 students. Courses taught:

  • Geometry Fundamentals
  • Chemistry Fundamentals
  • Electricity Fundamentals
  • Electrical Technology Lab
  • Electrical Circuits Lab
 
 
 
 
 
Escuela Politecnica Nacional
Laboratory Assistant
September 2014 – August 2015 Quito-Ecuador

Instructed undergraduate courses to groups of up to 50 students. Courses taught:

  • Geometry Fundamentals
  • Chemistry Fundamentals
  • Electricity Fundamentals
  • Electrical Technology Lab
  • Electrical Circuits Lab

Awards

Best Paper Award
Workshop on Effective Representations, Abstractions, and Priors for Robot Learning
Presidential Fellowship
Fellowship to support graduate students to continue their research and complete their dissertations.
Fulbright Faculty Development Scholarship
Scholarship provided to pursue graduate education in the USA.
Workshop Travel Award
Scholarship to attend coursework in Photovoltaic Solar Energy at TU Delft.

Projects

*
Template model inspired task space learning for robust bipedal locomotion
A hierarchical framework for bipedal locomotion that combines an RL-based high-level (HL) planner policy for the online generation of task space commands with a model-based low-level (LL) controller to track the desired task space trajectories.
Template model inspired task space learning for robust bipedal locomotion
Reinforcement Learning-Based Cascade Motion Policy Design for Robust 3D Bipedal Locomotion
A novel model-free reinforcement learning (RL) framework to design cascade feedback control policies for 3D bipedal locomotion.
Reinforcement Learning-Based Cascade Motion Policy Design for Robust 3D Bipedal Locomotion
Simulation of 2R planar robot using Gazebo and ROS
A 2R planar robot is simulated using Gazebo and ROS considering different values of height and mass for each robot’s link and different initial position of the joints. The dynamic behavior of the robot and its response to gravity is analyzed for each different setup of kinematic and dynamic conditions.
Simulation of 2R planar robot using Gazebo and ROS
Trajectory generation for quadruped robot leg
Implementation of trajectory generation for a prototype leg using inverse kinematics, cubic splines and Arduino.
Trajectory generation for quadruped robot leg

Recent Publications

Quickly discover relevant content by filtering publications.
(2023). Data-Driven Latent Space Representation for Robust Bipedal Locomotion Learning. Preprint.

Cite Video arXiv

(2023). Template Model Inspired Task Space Learning for Robust Bipedal Locomotion. In 2023 International Conference on Intelligent Robots and Systems (IROS).

Cite Video arXiv

(2023). MELP: Model Embedded Linear Policies for Robust Bipedal Hopping. In 2023 International Conference on Intelligent Robots and Systems (IROS).

Cite

(2023). Safe Bipedal Path Planning via Control Barrier Functions for Polynomial Shape Obstacles Estimated Using Logistic Regression. In 2023 International Conference on Robotics and Automation (ICRA).

Cite Video DOI arXiv