Overview
What this challenge is about.
You receive 4 hours of logged trajectories from the existing controller (joint positions, target poses, miss/success labels) and read/write access to the controller config (YAML). Tune the motion parameters using a structured search (start with one-at-a-time, then small grid). Validate on a held-out hour of logs by replaying targets in simulation. Success is miss rate cut by half on the held-out hour at unchanged or better cycle time. Deliver the tuned config, a tuning notebook, and a 1-page line-lead memo.
The Brief
What you'll do, and what you'll demonstrate.
Halve the missed-pick rate on a cosmetics co-packing line without sacrificing cycle time.
Earning criteria — what you'll demonstrate
- Tune a real robot controller's motion profile against measured outcomes
- Run a structured parameter search without overfitting to training data
- Validate controller changes via offline replay
- Communicate engineering changes to a non-engineer line lead
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.
AI Engineer
Tuning a real robot controller against production logs is everyday AI-engineer work in manufacturing-robotics teams; this challenge gives the student a concrete shipped change to point at.
This challenge sharpens
- motion-control
- python
- evaluation
Machine Learning Engineer
Structured search with held-out validation and Pareto reporting is the same discipline MLEs use when tuning models or pipelines.
This challenge sharpens
- trajectory-tuning
- evaluation
- python
Applied AI Scientist
Translating an engineering change into a line-lead memo is the applied-AI scientist's communication muscle at industrial-robotics companies.
This challenge sharpens
- motion-control
- evaluation
- trajectory-tuning