Prototype a Computer-Vision QA Tool for a Robotics Manufacturer
Overview
What this challenge is about.
As a 4-person team, build: (1) a labeling pipeline on around 2,000 component images (Label Studio is fine); (2) a transfer-learned classifier or a small segmentation model that flags surface defects; (3) a packaged inference service (FastAPI plus a Docker container) running on a small workstation; (4) a line-operator UI (web is fine) that shows the image plus pass/fail plus a confidence bar; (5) an evaluation report on a held-out 200-image set with precision, recall, and confusion matrix. Deliver a 6-page engineering writeup and a 25-minute on-site-style demo.
The Brief
What you'll do, and what you'll demonstrate.
Ship a working computer-vision defect-detection prototype with packaging plus an operator UI for a single production line.
Earning criteria — what you'll demonstrate
- Curate and label a real industrial CV dataset
- Train and evaluate a transfer-learned defect classifier
- Package a CV model into a deployable inference service
- Design a UI a line operator can use without training
Program Fit
Where this fits in your program.
Sharpens the same skills your degree expects you to demonstrate.
Skills
Skills you'll demonstrate.
Each one shows up on your verified credential.
Careers
Roles this prepares you for.
Real titles. Real skill bridges. Pick the one closest to your trajectory.
Career paths this builds toward
Canonical rolesComputer Vision Engineer
End-to-end CV ship across dataset, model, packaging, and UI is the literal CV engineer job at any robotics or manufacturing-AI shop.
This challenge sharpens
- computer-vision
- transfer-learning
- model-deployment
Machine Learning Engineer
Packaging a model into a reproducible deployable service is the MLE's daily craft.
This challenge sharpens
- model-deployment
- python
- transfer-learning
AI Engineer
Shipping a glued-together AI feature with operator-grade UI is the AI engineer's bread and butter in industrial-AI startups.
This challenge sharpens
- model-deployment
- user-interface-design
- team-collaboration