Engineering deep-dives for the people building Physical AI.
Eleven posts. Navigation, drive geometry, VLA++ multimodal fusion, Jetson Orin edge inference, and the data systems behind production robotics. Code, configs, and benchmarks where they matter.
The Open-Source Motor Design Stack: From FEMM to FOC Firmware
A working engineer's tour of the free tools, reference designs, and community projects you can actually use to design, simulate, build, and control electric motors for robots and EVs — without a single proprietary license.
Building a Bot Economically: Mobile Robot Bases vs Powered Wheelchairs
If you're building a service robot, the cheapest path to a reliable mobile base is rarely the one labeled 'mobile base.' A walk through Clearpath, AgileX, Hello Robot, and Intermode pricing — then a side-by-side with the powered wheelchair industry that already solved most of the same problems for less.
Drive Geometries for Car-Cleaning Environments: Ackermann, Differential, and Holonomic Under Real Navigation Constraints
This robot never goes inside the cabin. It navigates between vehicles, reaches different open doors from outside, and has to survive the aisle widths, floor conditions, and cost limits of real cleaning operations.