Quick answer first
LAMAR, a large jewelry manufacturer in Dubai serving Middle East retail and export markets, deployed a combined Cylanco CXM-C18 + CXM-C20 line and reached a 50% improvement in fully automated casting efficiency measured across three consecutive months. The result came not from a single "magic" parameter but from standardizing the upstream-to-pour workflow, tightening handoff discipline between shifts, and converting tacit operator know-how into a written, repeatable process sequence. This case documents what LAMAR changed, why the two-machine pairing mattered, and which operational habits other Middle East workshops can port directly.
Case context
LAMAR operates a mid-to-high volume jewelry casting line primarily serving the Middle East retail market, with export exposure into Levant and East Africa. Their product mix leans toward filigree and fine-detail 18K and 22K pieces — category demands that traditionally require two different process regimes: a thin-section filigree regime where fill fidelity dominates, and a higher-throughput solid/hollow regime where cycle time dominates.
Before the Cylanco deployment, LAMAR ran a single-machine line that forced operators to switch thermal profiles, vacuum timing, and pressure settings each time the product family changed. This created two hidden costs: inter-shift inconsistency (the same product family producing different outcomes depending on which operator was on duty) and handoff friction (defect causes diagnosed on one shift rarely carried forward to the next shift in a way the next team could act on).
Why two machines instead of one
Rather than escalating to a larger single machine, LAMAR deployed CXM-C18 dedicated to filigree and fine-detail runs, paired with CXM-C20 handling higher-throughput solid and hollow production. The rationale was less about raw capacity and more about eliminating regime switching: each machine now operated inside a narrower parameter envelope it was tuned for, and operators were trained against one profile per machine rather than three or four profiles that had to be selected at every shift handoff.
This is a common architecture pattern for workshops that hit a specific wall: they don't need more metal per hour — they need fewer product transitions per machine. In LAMAR's case, the transition between filigree and solid runs on the old single-machine setup was consuming roughly 10–15% of shift time in parameter re-validation and the first-lot verification that follows any profile switch. The two-machine pairing reclaimed that time by making each machine a steady-state performer.
Before optimization — the starting baseline
- Shift outcomes fluctuated even when the same machine, alloy lot, and investment lot were used. The operators understood why intuitively but the knowledge never became written SOP, so the next shift re-learned it.
- Defect closure relied on individual operator experience. A specific lead operator could reliably diagnose a pinhole cluster in two lots; without him, the same cluster took five or six lots to close.
- Handoff data was incomplete. Shift logs captured what was cast, not how it was cast or what compromises were accepted. This meant recurring defects were re-diagnosed from scratch rather than looked up.
- Upstream-to-casting timing was not measured. The time between investment mixing, burnout release, and actual pour varied by 30–40% across the week, and nobody knew because it wasn't on the log sheet.
Execution strategy — what LAMAR changed
1. One approved process sequence per product family
LAMAR committed to a single validated sequence per product family — filigree, solid, hollow — and forbade silent variations. When an operator wanted to change a variable mid-run, that change had to be logged in a change-request line before the next lot, not after.
2. Frozen change scope during validation
During any validation window, only one variable was allowed to move per controlled lot. This is the single biggest productivity gain most workshops underestimate: multi-variable tuning creates an unreadable signal and guarantees repeat work.
3. Upstream aligned to casting window, not department convenience
LAMAR treated burnout, investment handoff, and alloy preparation as part of the casting window rather than as separate department schedules. A flask leaving burnout did not sit indefinitely; it moved to the pour within a measured release-to-cast window. This one change alone removed a major source of the overnight-to-morning variation the workshop had been chasing.
4. Lot-level defect coding reviewed same shift
Instead of end-of-week defect reviews, LAMAR ran a 10-minute lot-level review at the end of each shift, with every defect coded against a standing list. Close actions were committed before shift end. This kept institutional knowledge inside the team rather than inside individual heads.
5. Cross-machine workflow alignment
Because the two machines handled different product families, LAMAR built explicit handoff protocols for products that could run on either machine — for example, a mixed lot that included both filigree and solid pieces. The rule: if more than 15% of the lot was mixed geometry, the lot was split across machines rather than compromising either machine's profile.
Result pattern
The reported 50% improvement in fully automated casting efficiency is a composite metric that combines cycle time reduction, first-pass yield improvement, and reduced shift-to-shift variance. The team is careful not to attribute the gain to any single knob. What changed was execution discipline — the new line became easier to predict, easier to troubleshoot when it drifted, and easier to train new operators on because the written SOP carried the institutional knowledge instead of asking the next operator to re-learn it.
Secondary outcomes LAMAR reported informally: onboarding time for new operators dropped because the training target was now one documented sequence per machine instead of "shadow the lead operator until you understand it." Defect recurrence dropped because the defect-coded shift log made the pattern visible within days instead of weeks. And export-grade QA pass rate improved on thin-section filigree pieces destined for high-scrutiny Middle East retail.
Replication checklist — what other workshops can port
- Commit to one validated sequence per product family; forbid silent variations.
- During validation windows, move only one variable per controlled lot.
- Treat upstream preparation as part of the casting window, not a separate department schedule.
- Close defect coding at end of each shift, not at the end of the week.
- If product-family mix exceeds 15% per lot, split the lot rather than compromising any machine's profile.
- Consider a two-machine pairing when product transitions consume >10% of shift time on a single-machine line.
Related reading: Humid-Weekend 18K Master Troubleshooting Workflow, CXM-C20 Vacuum Pressure Vibration System, Emerald (India) CXM-C20 Case Study, Jewelry Casting FAQ.