Robotics engineer · hardware founder · WPI ’26
Antonio
Marzoratti
I build robots, hardware, and the software that brings them to life.
For the past five years, I’ve helped take PAWE, a wheelchair power-assist, from a garage prototype to a granted patent — designing the mechanics, electronics, and firmware along the way. Now I’m also building Create OS, an AI workspace for engineering.
Franklin, MA · graduating Dec 2026 · open to robotics, embedded & hardware roles
02 / Selected work
From a wheelchair attachment to robots that find their own way.
PAWE — detachable power-assist for manual wheelchairs
A portable attachment that adds powered mobility to a manual wheelchair. Explore the patented drive mechanism, custom electronics, and firmware behind it — and the testing that shaped the design.
Open case study →
Create OS — an AI engineering workspace
Turn an idea into a connected engineering project, from requirements and parts research to firmware and a test plan. Forge, the built-in AI engineer, keeps track of what has been checked and what still needs review.
Open case study →
Autonomy & vision — WPI robotics sequence
A mobile robot that maps an unfamiliar maze and finds its way home. A camera-guided arm that sorts objects by color. A competition robot designed from first principles. Built with ROS, Python, and MATLAB.
Open case study →
04 / A lesson from testing
The best design decision started with a failed test.
PAWE’s feedback controller worked well on the bench. On the chair, a gentle curve became a spin. The difference forced me to look beyond the tuning and question what the sensors were actually telling us.
The wheel sensors measured speed, but not direction. When a wheel reversed during a turn, the controller misread it and kept adding power. The test recorded 230° of heading change on a straight command.
I switched to direct joystick control with smoother acceleration and kept the sensors for monitoring. The result was more predictable behavior. The detailed write-up explains the diagnosis, the tradeoff, and the sensor changes needed to revisit feedback control.
05 / Recognition
Recognition for the work.
1st place, CEO Global Pitch Competition (2024) · $25,000 Kalenian Award, WPI (2024) · VentureWell E-Team, $20,000 (2025) · top-25 e-Fest finalist (2025).
