--------------------------------------------------------------- | About me | Research | Explorations | Random | ---------------------------------------------------------------
Building quadruped robots from scratch and teaching them to walk, think, and maybe even play fetch
What if we could build a robot dog that actually behaves like a dog? Not just walking and running, but understanding its environment, responding to commands, and maybe even developing personality?
This exploration started with a simple question: How hard could it be to build a quadruped robot?
Turns out, pretty hard. But also incredibly rewarding.
Building physical systems taught me to respect the gap between simulation and reality. Every joint has backlash, every sensor has noise, and gravity never takes a break.
The mathematics of how to make systems behave the way you want them to is both elegant and practical. PID controllers became my best friends.
Version 1 barely stood up. Version 2 could walk forward. Version 3 could turn. Each iteration taught me something new about locomotion.
Designing the frame, figuring out servo placement, and learning that 3D printing takes forever when you need precise parts.
Getting the robot to stand up without falling over. Inverse kinematics became my daily vocabulary.
Implementing gait patterns, dealing with balance, and celebrating every successful step like it was a moon landing.
Adding behaviors, responses to environment, and working toward that dog-like intelligence that makes robotics magical.