CATIA V5FANUC Teach PendantNAAMSGD&T17-4 PH S.S.Root Cause AnalysisIn-House Machining
What Shipped
Five workstreams on the Model 3 body-in-white line during my co-op. Three were built, installed and are running in production; two were designed and handed to fabrication. The status column says which is which.
A bowl-feeder installation blocked one of the two locating features on a robot's end-of-arm tool. The pick still worked, but the part was no longer fully constrained. I measured the existing hardware and surrounding geometry, then designed replacement locating hardware in CATIA V5 that restores full 6-DOF part constraint while clearing the new feeder.
2 locating-assembly variants covering 3 fixtures, built from standard NAAMS components plus custom brackets and a 17-4 PH stainless upper locator
Produced the manufacturing drawings and carried the design through machining, finishing and assembly
Physically installed the tooling on the line myself
25-part live-line validation with no issues, followed by production monitoring — the hardware is running in production
In a three-robot FANUC spot-weld sequence, one robot ran roughly 10 seconds longer than the other two and set the station cycle time. Working with two senior engineers, I reprogrammed an adjacent robot in teach mode to take over one weld and its approach and home points from the bottleneck robot, and supported the robot-zone handoff retiming.
Station cycle time cut from 75 s to 65 s (−13%)
Checked interference and path coordination, then validated at full production speed
Monitored production stability with maintenance after release — the change held
Chamfered EOAT Clamp — In Production
A separate clamp on the same line was letting parts slide past the second locating pin. The clamp's contact face was too thick to engage cleanly where weld beads sit on the part ends. I redesigned it with a weld-bead clearance chamfer, had it machined in-house and installed it. It is running in production.
OEM Fixture Clamps — Drawings Complete
An OEM three-clamp fixture system used at eight production locations needed replacement parts, quoted at about $88,000 with a 10-month lead time. No original CAD existed. I reverse-engineered the clamp systems from the physical parts, modelled them in CATIA V5 and produced GD&T drawings so they can be fabricated in-house.
Full reverse engineering from physical parts, toleranced for machinability
The immediate production need was covered by a bushing replacement on the existing units
Nothing was machined under this workstream. The $88K is the OEM quote the drawings give an alternative to — not a realized saving
TCP Calibration Fixture — In Fabrication
Designed a 4-part offline check fixture so stud-welding guns can be certified against one shared master setup (TCP drift up to 2.5 mm), using datum-controlled locating, NAAMS L-blocks and 3-way shim adjustability. Drawings released; parts in machining — not yet installed.
Lessons Learned
Prove it on the line before asking for it
On a production line, gut feel gets you nowhere. Every change — the cycle-time rebalance, the locating hardware, the clamp — had to be backed by what the line actually did: a live validation run, maintenance sign-off and production stability afterwards. Learning to prove a change before asking for it was as important as the mechanical design work.
Change management is a real engineering skill
Getting a change onto a live production line involves coordinating with operators, process engineers, and controls — not just submitting a drawing. Understanding how to communicate a change, validate it, and get sign-off was something I had to learn on the job.
Say which work shipped
Two of these five workstreams stopped at drawings, for good reasons — a bushing swap covered the immediate clamp need, and the TCP fixture was still in the machine shop when the co-op ended. Stating that plainly is what makes the three production claims believable.