SKU Changeover on Indian Packing Lines: Where the Hours Go

A SKU changeover costs you twice. First the obvious part: the line is stopped while you swap formats, reels, coding and recipes. Then the part nobody counts: the first hour after restart runs slow and throws more rejects while the filler re-centres and the sealer finds its temperature. On a line worth ₹50 a minute, a 90-minute changeover plus a soft first hour can quietly cost ₹6,000–7,000 — and a plant doing four or five changeovers a day loses the better part of a shift.

By the Mama Editorial Team. Short runs and many SKUs are the reality of the modern Indian packing hall — more flavours, more pack sizes, more private-label. That flexibility is sold one changeover at a time. This article shows where those hours actually go, works the rupees, and lists what cuts them.

Key takeaways

• Changeover loss has two halves: stopped time (the swap) and ramp-up loss (a slow, reject-heavy first hour). Most plants log only the first.
• The first hour is the expensive one — filler weights settle, sealer temperature stabilises, operators re-learn the format, and rejects spike before the line finds rhythm.
• Value a changeover in contribution margin per minute, not just labour. A stopped line loses the margin it would have earned.
• SMED is the proven method: convert internal setup (line stopped) to external (done while running), stage changeover kits, and sequence SKUs to minimise the size of each change.
• A camera times this honestly — stop, restart, and the slow ramp — but it cannot judge seal integrity or pack weight inside the pack.

Why a changeover is a downtime problem, not a maintenance one

On a packing line a changeover is a planned stop: you are converting the line from making SKU A to making SKU B. Film reel, pouch format, carton size, date/batch coding, product recipe, fill target, sealer settings — some or all of these change. While that is happening the line earns nothing, and in OEE terms it is a straight Availability loss.

The mistake is to treat the stop as the whole cost. The stop is only the part you can see on the clock. The rest is hidden in the shift that follows.

The two costs of a changeover

1. Stopped time — the swap. From last good pack of SKU A to first pack of SKU B. This is what most shift books record (if they record it at all).

2. Ramp-up loss — the first hour. After restart the line does not jump to full speed and full quality. Several things settle at once:

• The filler re-centres. Until the new target weight is dialled in, you either over-give (see our piece on product giveaway) or throw underweight rejects.
• The sealer finds temperature. A cold or over-hot seal bar means leakers and weak seals until it stabilises — a reject spike on a food pack.
• Coding and registration settle — misprints and off-centre cuts in the first minutes.
• Operators re-learn the format: the first SKU B of the day is always clumsier than the hundredth.

So the real cost is: stopped minutes × margin/min plus ramp-up minutes of lost speed and extra rejects. Leave out the second term and you under-count changeover loss by a third or more.

Worked example, in rupees

One line, illustrative numbers — substitute your own.

Stopped time: 90 min × ₹50/min = ₹4,500.

Ramp-up speed loss: running the first hour at 60% means losing ~40% of 60 min = ~24 min of output × ₹50 = ₹1,200.

Ramp-up rejects: 150 packs × ₹2 margin (plus scrapped material, ignored here to stay conservative) = ₹300.

Total per changeover ≈ ₹6,000. The stopped time alone would have told you ₹4,500 — a 25–30% under-count.

A plant running four changeovers a day on one line is losing the thick end of ₹70 lakh a year to SKU changes — most of it invisible because the ramp-up half never hits a report.

How to cut it: SMED, adapted for an Indian packing hall

The discipline for this is SMED — Single-Minute Exchange of Die, Shigeo Shingo's method for slashing setup time. The core idea travels straight to a packing line:

• Separate internal from external setup. Internal = must be done with the line stopped. External = can be done while the line still runs SKU A (fetching the next reel, pre-kitting cartons, pre-setting the coder). Most plants do far too much inside the stop. Moving even half of it outside is the single biggest win.
• Stage a changeover kit for each SKU-to-SKU transition: the right reel, change parts, cartons, labels, recipe card — pre-assembled on a trolley so nobody walks to the store mid-stop.
• Standardise settings as recipes. Fill target, sealer temperature, speed, coder layout stored per SKU and recalled, not re-found by trial. This is what kills most of the ramp-up reject spike.
• Sequence your SKUs to shrink the change. Run same-pack-size or same-film SKUs back-to-back so a "changeover" is a recipe recall, not a full format strip-down. Group by the biggest, slowest thing to change.
• Quick-release change parts instead of bolted ones, where the machine allows.
• Measure every changeover — start, end, and when the line reached stable full rate — so you know which transitions overrun and why. You cannot shorten what you do not time.

The same logic shows up across discrete manufacturing — see how it plays out with job changes in packaging and printing plants.

What a camera can — and cannot — time

Here is where a camera genuinely earns its place, and where it does not. Video is good at the time boundaries of a changeover: it sees the last pack of SKU A, the line going still, the crew working, the restart, and — importantly — the slow ramp, the minutes before the line is back at full rate. That is exactly the data most plants are missing: not "there was a changeover" but this changeover took 112 minutes and the line did not hit full speed for another 40, while the one before it took 70. It can also flag the honest difference between a crew working the changeover and a line sitting idle mid-changeover because someone went looking for a part — the external-setup failure SMED targets.

What a camera cannot do: judge whether a seal is sound, read the weight inside a pack, or certify print quality. Those are the checkweigher's, the leak-tester's and QA's job. A camera times the hours; it does not inspect the product.

That split is the honest pitch for Mama: watch the line from your existing cameras (or ones we place) and every morning send the owner, on WhatsApp, where the hours went — including each changeover timed start-to-stable and the ones that overran — with a rupee estimate. It will not tell you a seal failed; it will tell you the ₹ sitting in your changeovers. Send a short video of the line and we will return where the hours go, a camera plan, and a proposal.

How a plant head should start

• Time ten changeovers properly — to stable full rate, not just restart. The ramp half is usually the surprise.
• Split the task list into internal vs external. Move everything you can outside the stop.
• Build SKU recipes for fill, seal and coding to kill the reject spike.
• Sequence the schedule to minimise change magnitude, not just to please dispatch.
• Put a number on it — margin/min × stopped-plus-ramp minutes — so the shift floor sees the rupees, not just the clock.

FAQ

What is a SKU changeover on a packing line? Converting the line from one product/pack (SKU A) to another (SKU B) — swapping film, format, coding, fill target and sealer settings. The line earns nothing while it happens and runs below par for a while after.

Why is the first hour after a changeover so expensive? The filler is still re-centring its weight, the sealer is reaching stable temperature, coding and registration are settling, and operators are re-learning the format. Speed is low and rejects are high until it all settles — a loss most shift books never record.

How do I calculate the cost of a changeover? Margin per running minute × (stopped minutes + ramp-up minutes of lost output) + the cost of ramp-up rejects. Value the line in contribution margin, not just labour.

What is SMED and does it apply to packing lines? SMED (Single-Minute Exchange of Die) is Shigeo Shingo's setup-reduction method. Yes — its core move, converting internal setup to external and standardising settings, cuts both the stop and the ramp on a packing line.

Can a camera measure changeover time? Yes, for the time boundaries — stop, restart and the slow ramp to full rate — and it can spot a line sitting idle mid-changeover. It cannot judge seal integrity, pack weight or print quality; those stay with the checkweigher and QA.