SMED (Single-Minute Exchange of Die)

SMED (short for Single-Minute Exchange of Die, meaning a die change completed in single-digit minutes) is a lean manufacturing methodology whose goal is to drastically reduce setup time: the span during which a machine sits idle while the mold, die, product format or tooling is changed, from the last good part of one product to the first good part of the next. The original target, set by the Japanese engineer Shigeo Shingo, is to complete that changeover in under ten minutes (a single-digit number).

The central idea is to distinguish between internal tasks (those that can only be done with the machine stopped, such as loosening and mounting the mold) and external tasks (those that can be prepared while the machine is still running, such as bringing tools over, preheating or having the material ready). SMED seeks to convert as many internal tasks as possible into external ones and then simplify and standardize what remains. It is one of the techniques that organizes a plant's flow within the lean manufacturing approach, and it applies naturally to lines that produce short, varied batches, a common pattern in the Manufacturing industry in Argentina and Latin America.

Why SMED matters

In any plant that makes more than one product on the same line, changeover time is a huge hidden cost. While the machine is stopped for a changeover it produces nothing, but it keeps consuming labor, energy and, above all, available capacity. The longer the setup, the more tempting it is to run large batches to "amortize" the changeover, which inflates inventory, stretches lead times and hides defects. SMED breaks that cycle: if switching products takes minutes instead of hours, the plant can run smaller batches, respond faster to real demand and cut stock without losing efficiency.

The impact shows up directly in OEE (overall equipment effectiveness): long changeovers hit the availability component, so shortening them raises OEE without buying new machines. That is why SMED is usually one of the first improvement projects in a lean transformation: low implementation cost and fast, measurable results.

How it works: the classic stages

Shingo described a staged method that is still the practical reference:

• Stage 0, diagnosis: the entire changeover is filmed and every step is timed. In most plants that have never measured it, internal and external tasks are mixed together with no order.
• Stage 1, separate internal from external: you identify the tasks that could actually be done with the machine running (finding tools, preheating molds, preparing material) and take them out of the downtime. Just sorting this out usually cuts a good part of the changeover time.
• Stage 2, convert internal into external: tasks are redesigned so they no longer require the machine to be stopped. Typical examples: preheating a mold separately instead of heating it inside the press, or having a second set of tools already mounted.
• Stage 3, simplify and standardize: adjustments and bolts are eliminated, quick fasteners, guides and locators are used, and the sequence is standardized so any operator performs it the same way.

A concrete example from the region

Think of a plastic packaging plant in Greater Buenos Aires that injection-molds caps of different diameters on the same machine. Before SMED, changing the mold took 90 minutes: the operator shut down the press, only then went to get the crane and the tools, waited for the new mold to reach temperature inside the machine and adjusted by trial and error. With a SMED project, the team left the next mold preheated and ready next to the press (an external task), replaced the bolts with quick-release clamps and standardized the parameters. The changeover dropped to 12 minutes: the plant could make smaller batches, handle urgent orders and raise its availability without investing in a second injection molding machine. This kind of improvement goes hand in hand with practices such as Kanban, so the production flow keeps pace with the smaller batches.

Common mistakes

• Thinking SMED means buying expensive equipment: most of the progress comes from reorganizing tasks, not from spending.
• Skipping the filming and measuring stage: without data, the team argues over opinions instead of attacking the steps that really consume time.
• Doing the project once and not standardizing: if the improvement does not end up in a work instruction and a standardized setup, the time creeps back up within a few weeks.
• Confusing changeover time with machine downtime caused by breakdowns: SMED targets the product changeover, not breakdowns (that is a separate effort, tied to maintenance).

SMED compared with other plant metrics and techniques

In short, SMED is a flexibility lever: it turns changeovers that took hours into changeovers that take minutes, enables production in small batches and raises a plant's effective capacity without heavy investment. It is one of the lean tools with the best effort-to-return ratio, which is why it shows up so early in almost every continuous improvement program in manufacturing.