Production Data Control

That was the gap CDMS didn't close, because CDMS ended where distribution handed fabric panels into sewing. What happened after that, piece by piece, line by line, all the way to folding, stayed manual, and manual meant it was easy to misreport.

After CDMS shipped, I was asked to lead the next phase: Production Data Control (PDC), built to serve sewing, industrial engineering, and QC directly. By this point I wasn't working as a junior on someone else's project anymore, I was leading this one, still alongside the same OE counterpart, but now driving the system side myself.

PDC had to work across three factories that didn't share the same physical setup. One ran fully manual sewing, no conveyor at all. Another ran on INA hangerline, an RFID-tracked overhead system that routes garment pieces automatically between workstations. The third used a simpler, non-RFID hanger system, automated movement without the tracking layer. Three different ways of physically moving a garment through a line meant three different ways of capturing what actually happened at each one.

When output is self-reported at each station, with no independent checkpoint, there's no way to tell "on track" from "marked on track." That gap between reported progress and actual progress was the first thing PDC needed to close, not through more paperwork, but by making every checkpoint, loading, sewing, QC, folding, capture data at the moment it actually happened, not at the moment someone chose to report it.

This project sat differently on my plate than CDMS did. On CDMS, the cross-factory meetings and standardization process were driven by my OE counterpart, with me supervising the technical build alongside him; here, I initiated and ran the standardization process myself, on top of the same development supervision. Less junior, more ownership, same system-side role at the core.

The bigger technical challenge was that "capture data at the moment it happens" meant something different at each of the three factories. One ran fully manual, no automated tracking at all. One ran INA hangerline, RFID tags on every hanger, scanned automatically as garments moved station to station. One ran a simpler hanger system, automated movement but no RFID layer underneath it. Three factories, three different physical realities, and PDC had to make all three report through the same system without pretending they worked the same way.

Instead of a direct, live integration, data from the INA hangerline landed in a staging database first, and PDC pulled from that on its own schedule. It added a small delay, but it meant a vendor-side failure stayed a vendor-side problem instead of taking down production tracking with it.

The other two factories, the fully manual one and the one running a non-RFID hanger system, needed a different fix: two barcode scanners, one at loading, one at folding, with output per piece captured directly at the sewing station itself, a button press the moment a piece was finished, not self-reported at the end of a shift. Neither setup had automatic tracking, so the same manual-capture approach covered both.

Access to PDC was role-based from the start, with correction rights limited to production admins rather than open to whoever was at a station. Whatever got captured at the moment of completion stayed the record; fixing it required going through someone accountable for the change, not just re-entering a number.

All of it ran on a monthly cross-factory sync across sewing, IE, and QC at all three plants, plus ad hoc meetings when something couldn't wait.

This time, initiating and running those standardization meetings was mine to own, not something I stepped into after someone else had already set the agenda. Reporting to management followed the same shift: I led those conversations directly, bringing in a senior developer when a question went deep enough into the hangerline integration that it needed someone closer to the build to answer it well.

Output could be monitored at any level: by line, by station, by individual operator, by piece.

Folding felt the difference most directly. Instead of finding out about a shortfall only when a full batch failed to arrive, folding could see exactly which piece was missing and which station it was still sitting at, and follow up there directly instead of chasing it blind.

PDC now runs across all three factories, with integration work still ongoing, most recently a planned move to RFID operator ID cards, so operators can tap in, out, and for breaks at their own station instead of manual entry. I moved on from Apparel One before that phase shipped, but the system itself is still active and still being built on.