Operations
How to increase jewellery production capacity without rebuilding your factory
Buying a faster machine rarely raises output on its own. Here is how to find the actual constraint on your floor — and what to do about it in what order.
· · 10 min read · By Semsons Tech
A factory owner rings us and says: "I need a faster setting machine." Perhaps a third of the time, they do. The rest of the time, the setting bench is not what is holding them back — and if we simply sold them the machine they asked for, their output would barely move.
This is how to work out which situation you are in.
The rule that governs every factory
Your output is set by your slowest stage. Not your average speed — your slowest stage. Every other stage can only ever run as fast as the constraint allows, whatever its rated capacity.
Adding capacity anywhere except the constraint produces more work-in-progress, not more finished pieces. You have not increased output. You have increased the size of the queue.
This is why "buy a faster machine" so often disappoints. It is the right instinct applied at the wrong point in the line.
Finding your actual constraint
You do not need software for this. Walk the floor at your busiest point in the month — not a quiet week, because constraints hide when demand is low — and look for two things:
- Where is work waiting? Inventory piles up immediately before the constraint. Trays queued at the setting bench, flasks waiting for burnout, baskets stacked at polishing.
- Where are people idle? Stages after the constraint get starved. If your casting operator is genuinely idle for hours, casting is not your problem — whatever feeds it is.
The stage with work piled in front of it and idle capacity behind it is your constraint. That is where the next rupee of capital belongs.
The usual suspects, by stage
| Stage | Typical symptom | Usual fix |
|---|---|---|
| 3 — Wax injection | Rejects appear downstream at setting; wax varies batch to batch | Vacuum wax injector, temperature discipline |
| 4 — Diamond setting | Setters on permanent overtime; trays queued; delivery dates slipping | Automatic setting machine |
| 5 — Investment & burnout | Flasks waiting; furnace cycles are the daily schedule driver | Additional burnout furnace capacity |
| 6 — Melting & casting | Casting runs late into the evening; slow melt cycles | IGBT induction melting |
| 7 & 9 — Polishing | Finished-but-unpolished stock accumulating | Bulk magnetic or vibratory finishing |
| 10 — Measurement & QC | Goods held pending verification | XRF gold analyzer |
When setting genuinely is the constraint
It often is, and for a structural reason: setting is the most labour-intensive stage on the floor, and skilled setters are the hardest people to recruit. You cannot buy your way out with hiring at short notice.
The catalogue figures are worth stating plainly. One Mini Pro is claimed to do approximately the job of 10 workers at up to 10,000 stones per hour. The ST-3D 1000 and ST-HDSSM are each claimed at approximately 15 workers, rated at 18,000–20,000 diamonds per hour, holding ±0.02 mm placement accuracy throughout.
But before ordering, check the two stages either side.
Check upstream first
A setting machine placing stones to 0.02 mm into inconsistent wax will produce precise errors. If your wax injection is not under control, fix that first — see why wax quality decides your setting machine's real output. This is the single most common reason a new setting machine underdelivers.
Then check downstream
If you triple setting output, everything after it must absorb the flow. Burnout furnace cycles, casting capacity and polishing throughput all become the new constraint the moment setting stops being one. Plan for that rather than discovering it.
Capacity you can add without buying machines
Before capital, exhaust the free options — they are often worth 10–20% and they cost nothing:
- Batch by design family. Every changeover costs setup time. Grouping similar designs reduces them.
- Stagger shifts around the constraint. Keep the bottleneck stage running through breaks. It is the only stage where an idle minute is genuinely lost forever.
- Move quality checks earlier. Catching a wax fault at stage three costs a wax. Catching it at stage nine costs everything you spent since.
- Fix the compressed air. If air pressure sags under load, pneumatic equipment underperforms in ways that look like machine faults.
- Clean the dust. Polishing dust in precision machinery causes downtime that nobody attributes to housekeeping.
A sequence that works
- Identify the constraint by walking the floor at peak.
- Exhaust the free fixes at that stage.
- Verify the stage upstream can feed a faster constraint.
- Verify the stage downstream can absorb the output.
- Invest at the constraint.
- Let the floor stabilise for a few weeks.
- Find the new constraint — there will be one — and repeat.
This is slower than re-equipping everything at once, and it is why factories that phase their upgrades tend to end up with better floors than those that spend big in one go. Each step is partly funded by the gain from the last.
We would rather tell you not to buy
If you send us your production figures and we think setting is not your constraint, we will say so — and point you at the stage that is, even when that means a smaller order or none at all. A customer whose output actually rises comes back for the next nine stages. One who bought the wrong machine does not.
Send us your production figures, or read the full ten-stage factory workflow to see how the stages fit together.
Answers
Related questions
How do I find the bottleneck in my jewellery factory?
Walk the floor and look for where work is waiting. Inventory piles up immediately before the constraint and starves the stages after it. If trays of wax are queued at the setting bench while casting sits idle for half the day, setting is your constraint. Do this at your busiest time of the month, not a quiet week — constraints hide when demand is low.
Will buying a faster machine increase my output?
Only if that machine is at the constraint. Capacity added anywhere else produces more work-in-progress, not more finished pieces — it moves the queue rather than removing it. This is why we ask about your whole workflow before recommending a machine, and why we sometimes advise customers not to buy the machine they asked about.
Should I upgrade one stage at a time or all at once?
One at a time, in constraint order. Fix the bottleneck, let the floor stabilise for a few weeks, then find the new bottleneck — because there will be one, somewhere else. Phasing spreads the capital, and each step is funded partly by the gain from the last.