01
Legged Locomotion
Developing coordinated 12-joint motion, standing stabilization, pose transitions, recovery, and gait behavior for execution on physical quadrupeds.
Dynamics · Stability · Gait Design Independent Robotics Research · Dallas, Texas
Building more capable
legged machines.
Quad Research Lab is an independent robotics research laboratory focused on legged locomotion, reinforcement learning, robotic mechanisms, proprioceptive actuation, and the transfer of learned control from simulation to physical hardware.
ACTUATIONPROPRIOCEPTIVE
LOCOMOTIONRL POLICY
PIPELINESIM → REAL
Studying the complete stack required for dynamic legged robotics—from physical architecture and actuator integration to learned locomotion and robust deployment on real machines.
Current quadruped research platform
Physical development is centered on a custom quadruped platform used for actuator integration, low-level control, learned locomotion, and repeated simulation-to-hardware testing.
01 / RESEARCH
Core research areas
Research is organized around the systems that determine how a legged robot moves, learns, and survives outside simulation.
02
Reinforcement Learning
Training neural locomotion policies with MuJoCo, MJX, JAX, PPO, and PyTorch-based workflows, with emphasis on policy-only control and sim-to-real transfer.
MuJoCo · MJX · PPO · Policy Learning03
Proprioceptive Actuation
Characterizing and integrating actuators with position, velocity, torque, and feedback control for safe, responsive learned locomotion on real hardware.
Actuation · Feedback · Embedded Control04
Robotic System Design
Iterating mechanical linkages, structures, cable routing, electronics, and power architecture while keeping the physical robot and simulation model closely aligned.
Mechanisms · Electronics · Integration
02 / ABOUT
Independent research.
Independent research.
Built around the machine.
Quad Research Lab is an independent robotics research laboratory in Dallas, Texas, founded by engineer Han Yildirim. The lab exists to advance the physical capability and intelligence of legged robots through tightly coupled work in mechanics, electronics, control, learning, and experimentation.
Current work centers on building and validating custom quadruped systems, developing proprioceptive actuation and control pipelines, and pushing learned locomotion toward reliable execution on real machines.
04
Primary Research Directions
∞
Iteration between design, simulation, and hardware
03 / CONTACT
Research, technical collaboration,
Research, technical collaboration,
and supplier inquiries.
EMAIL
han.yildirim@quadresearchlab.com
Quad Research Lab · Dallas, Texas, USA