When a line runs slow, the instinct is to speed up the machine. But the most volatile lines in our study weren't slow because their machines were slow. They were slow because they never stopped stopping. That fragmentation carries a measurable, non-negotiable penalty.
A line isn't hitting target. The obvious move is to push the machine harder, or to budget for a faster one. It feels like the right lever, and sometimes it is. But across 62 lines we found a pattern that says: check the rhythm first, because you may be about to pay for speed you can't use.
We split lines by how fragmented their running was, measured as time between interruptions. The most stable group posted an average Performance score of 81.2%. The most volatile group, stopping and starting constantly, dropped to 68.7%.
That 12.54% gap is a tax on fragmentation. The volatile lines weren't running slower machines, they were running the same machines in a state of permanent interruption, so those machines spent their lives decelerating and ramping back up instead of producing.
Performance, in OEE terms, is actual rate against ideal rate. On a fragmented line the losses don't come from a low ideal rate, they come from never sustaining it. Add a faster machine and you simply arrive at the next micro-stop a little quicker. The constraint isn't the equipment's ceiling; it's the floor of constant interruption underneath it.
This is why the Speed Tax is so expensive: it's invisible to a capacity plan built on nameplate speeds, and immune to the capital that plan usually recommends.
The 12.54% of Performance maps to roughly 4.89% of total OEE, real capacity, sitting idle. Recovering it means attacking the highest-frequency stop causes to lengthen the time between interruptions, not accelerating the machine. It's cheaper than capital, and it's usually faster.
The Speed Tax is one of the core findings of the Interruption Whitepaper. To see whether a given line is paying it, start with its time between interruptions and its operational-health tier.
It's the performance penalty a line pays purely for process fragmentation, the constant stopping and restarting that prevents it from ever running at a steady rate. In our study it measured 12.54% between the most stable and most volatile cohorts.
Because the machine's top speed is rarely the limiting factor on a volatile line. Time is lost decelerating, sitting idle and ramping back up between micro-stops. A faster machine just reaches the next stop sooner.
The 12.54% performance gap corresponds to roughly 4.89% of total OEE. That's recoverable capacity, realised by stabilising the line's rhythm rather than by capital spend on equipment.