Project STEP (GY)

  • Designing and building humanoid robots from ground up to automate daily tasks.

Connect

Robot looking thing on two legs

Humanoid robots have long captured the imagination of roboticists, representing the pinnacle of advanced robotics and AI. Despite the intense research interest in this area, the high cost of research remains a significant barrier. We aim to tackle this challenge by offering students hands-on opportunities to build and work with humanoid robots. Our mission is to make humanoid robotics research more accessible through an open-source, low-cost platform that empowers students to engage in cutting-edge robotics. By leveraging recent advances in reinforcement learning, deep learning, additive manufacturing, and actuators, we strive to create an inclusive and innovative learning environment.

 

Goal

Our goal is to develop an accessible bipedal robot research platform for humanoid robotics. Our focus areas include:

  • Lower-body development: Manufacturing, gait design, and real-world deployment.
  • Perception systems: Enhancing the robot’s ability to interact with its environment.

Upper-body & whole-body manipulation: Advancing the dexterity and functionality of bipedal robots.

 

Click here for video

Methods & Technologies

  • Reinforcement Learning (PPO, GRPO, etc.)
  • Optimal Control Theory (SQP, MPC)
  • Simulation Platforms (Isaac Gym, MuJoCo)
  • CAN bus protocol implementation
  • Firmware development & Embedded systems
  • CUDA and GPU programming
  • Sensor integration and signal processing
  • Actuator control & wire harness design
  • 3D CAD modeling (SolidWorks, OnShape, Fusion 360)
  • Finite Element Analysis (FEA, MATLAB)
  • Additive & subtractive manufacturing (3D printing, CNC milling, laser cutting)
  • Minimum Viable Product (MVP) prototyping
  • Public relations and outreach
  • Fundraising and sponsorship management
  • Website design and content creation

Areas of Interest

  • Computer Science & Algorithm Design
  • Electrical Engineering
  • Mechanical Engineering
  • Business and Marketing

Subteams

  • Computer Science (CS) Team: Robotics software, control algorithms, and reinforcement learning.
  • Electrical Engineering (EE) Team: Firmware, communication protocols, and tuning proportional (KP) and derivative (KD) gains for sim-to-real transfer.
  • Mechanical Engineering (ME) Team: Hardware design, testing, and manufacturing.
  • Operations Team: Purchasing, marketing, website management, fundraising, and outreach.

Partners

  • Applied Dynamics and Optimization Laboratory

Faculty Advisor