AutoPatt
2024
Automated system using image processing and motor control to achieve sub-micrometer precision in photolithography for Intel.
Overview
AutoPatt is a hands-on prototype that demonstrates how low-cost hardware and open-source vision algorithms can perform precision alignment for photolithography experiments. Built during HackOHIO 2023, the system integrates image capture, geometric estimation, and closed-loop motor control to align micro-scale patterns reliably.
"We proved that careful software plus modest hardware can reach precision people often assume requires expensive lab equipment." — Kuldeep
Key Features
- •Automated image capture with OpenCV-based displacement and rotation estimation.
- •Closed-loop motor control using Arduino for sub-micrometer adjustments.
- •Web-based UI to visualize alignment status and trigger calibration steps.
Applications
- •Educational demonstrations for microfabrication and controls courses.
- •Prototype alignment system for small-scale research or hobbyist use.
- •Proof-of-concept for low-cost tooling in early-stage semiconductor workflows.
Tech Stack
Key Achievements
- •Achieved repeatable sub-micrometer alignment on a DIY stage.
- •Presented a working prototype at HackOHIO with live demos.
- •Hosted a public project site showcasing design, code, and technique (moreUrl).
Impact
AutoPatt reframes the accessibility of precision fabrication: by lowering the cost and technical barrier, educators and small research teams can teach and iterate on photolithography concepts without expensive cleanroom access. The project inspired peers to explore hybrid hardware/software solutions for precision engineering.