🔀 Product Sequencing

One line, several products, and a changeover every time you switch. The order you run them in is a real decision — and there is no single right answer. Set up your line below and see three different plans: the one that costs least, the one that finishes soonest, and the one that keeps you in stock. A product can appear more than once: sometimes you must come back for a second run to stop something else running dry.

1 · The line

How many products?

Time unit

2 · Changeovers

What it costs and how long it takes to switch the line from one product to another. Direction matters — going from a dark colour to a light one is usually a longer clean-down than the reverse.

Running two batches of the same product back to back needs no changeover — it is free, in both time and money.

3 · Three sequences, three priorities

4 · The trade-off, side by side

Every sequence scored on every measure. ★ marks the best in each column — if the stars scatter across rows, you are looking at a genuine trade-off. If one row takes them all, you got lucky.

5 · What that plan actually looks like

Each line is one product's inventory. It slopes down as demand eats into it and jumps up when a batch comes off the line. Anything below the zero line is a stockout — demand you could not serve.

Why the three answers disagree

Total production time is fixed — the same batches have to run either way. So sequencing only moves three things around, and they pull in different directions.

Total cost = changeover cost + holding cost + stockout cost Cost-optimal sequencing chains products in the order that makes changeovers cheap — often grouping all batches of a product together, because a repeat run is free. That is exactly what builds inventory you then have to carry, and what leaves the last product waiting.

Average completion time = mean of "when was each product finished?" Finishing soonest rewards putting short runs first — everyone behind a long campaign waits for it. Minimising the makespan (when the last unit is done) is a different question again, and it only ever squeezes total changeover time.

Stockout exposure = total time any product sits below zero This one ignores money entirely and just asks who is about to run dry. It front-loads whatever has the shortest runway — current inventory ÷ demand rate — even if getting there means an expensive changeover, and even if it means splitting a product into two visits.

The model, stated plainly

One line. It makes one product at a time, and everything is made in whole batches.
A product needing more than one batch can be split up and revisited later in the sequence — that is the 2 → 1 → 3 → 2 pattern.
A batch's output lands in stock when the run finishes, not while it is running.
Demand keeps draining stock the whole time, including during changeovers.
The line starts clean — the first product in the sequence pays no changeover.
A part-full final batch still occupies the line for a full batch lead time.
Unmet demand is backordered, not lost: inventory goes negative and you pay the stockout rate per unit per period until you catch up.
The horizon ends when the last batch finishes; nothing after that is scored.

Holding cost per unit per period is the same idea as on the Carrying Cost page — if you are not sure what to type, build the rate there first. And if your demand figures are shakier than a single average suggests, that scatter is what Demand Variability is about.