01. Challenge
Torque-to-failure (TTF) testing for a specific next-gen vehicle required multiple tests to be
done in-line. The original method relied on manual point picking to find peak prevailing, snug,
and yield points from the data traces. This was slow and inefficient for high-volume testing.
02. Action
Algorithm Development: I built a Python-based tool that processes
angle-to-torque graph traces. It automatically determines the peak prevailing, snug, and yield
points based on user-defined parameters using a specific reference point scanning logic.
Reporting: Automated the workflow so that after point picking, a PDF report is
immediately exported, allowing engineers to review data summaries at a glance.
Deployment: Designed the tool to be compatible with micro-control tools used in
Gigafactories at Reno, Berlin, and Shanghai.
03. Impact
- Reduced analysis time by 30+ minutes per joint.
- Successfully deployed the tool to global teams in Reno, Berlin, and Shanghai, standardizing
the TTF reporting process.