Dain Kim
Real hardware. Documented tradeoffs. Every project here was built, debugged, and written up — so you can see how it works and why it was done that way.
Built and shipped
Brushless motor controller
Vision-guided pick-and-place arm
Lidar-based autonomous platform
Custom FOC controller from schematic to firmware. Tuned current loops on bench hardware; documented every iteration of the timing fixes that finally made it stable.
Ground-up SLAM implementation on a differential-drive chassis. Chose scan-matching over a particle filter; documented why and what the map quality difference looked like.
Six-axis arm with wrist-mounted depth camera. Solved the hand-eye calibration from first principles; wrote up the tradeoffs between extrinsic methods.
Why I stopped trusting my PID simulation
The reasoning behind the build
Hardware introduced latency the sim never modeled. Here is what changed when I measured it directly and how I adjusted the loop gains accordingly.
Each post covers one specific decision: which approach, which constraint forced the choice, and what the data showed. Control systems, sensor fusion, mechanical design — written for engineers who want the details.
Choosing between EKF and UKF on constrained hardware
The UKF converged faster but the sigma-point cost was too high at 100 Hz. A breakdown of the compute budget and why EKF won on this platform.
Depth is in the details
Every project links to the post that explains the approach. Start with the work, then follow the reasoning into the blog.