Assembly Line Balancing and Bottleneck Analysis

A line that can produce twenty boards an hour produces the number its slowest station allows, and every station upstream of it has to wait. Improving any other station adds work in progress rather than output, which is why the bottleneck is the only place an improvement pays.

Finding the Bottleneck

The bottleneck is the station with the longest effective cycle time, and the effective figure includes the downtime, the changeovers and the rework that lands on it. A machine that is fast when it runs and stops often can be slower than one that runs steadily.

Measuring it requires the uptime and the cycle time together over a period rather than a stopwatch reading on a good hour. Our first pass yield notes describe how the rework load is attributed to a station.

Why Local Improvements Fail

Improving a station that is not the constraint increases the queue in front of the constraint and changes nothing about the output. It also hides the constraint behind a buffer, which makes it harder to find.

The exception is an improvement that reduces the work the constraint has to do, which is the reason quality work at an upstream station can raise output. Reducing what the bottleneck reworks is the same as making it faster. Our placement capability notes describe one place where that pays.

The effective cycle time of a station is the run time plus the setup time divided by the batch size. A large batch hides a long setup, and a small batch makes the same setup dominant.

That is why the same line produces different throughputs for different products without anything about the machines changing. The bottleneck moves with the product mix, and the answer to a question about capacity has to name the product.

A buffer in front of the constraint protects it from short stops upstream. A buffer is not a solution to a constraint, and a buffer that is allowed to grow beyond its purpose turns a short stop into a long one.

The size of a buffer is set from the variability of the station upstream and the cost of the constraint stopping. It is a calculated figure rather than a space that happens to be available. Our floor zoning notes describe how that space is arranged.

A station that reworks its own defects has a capacity that is partly consumed by the rework. That consumption is invisible in a production plan that counts only new work.

Where rework is measured and attributed, the constraint often turns out to be a station with an adequate cycle time and a rework load. That is a quality problem presenting as a capacity problem.

The order that works is to measure, to protect the constraint, to reduce what it has to do, and only then to raise its rate. Raising the rate of a constraint that is reworking is moving the problem rather than solving it.

The measurements that support the sequence are the cycle time, the uptime, the rework load and the queue in front of each station. Four figures per station over a shift are enough. Our board quality notes describe the standard the rework is measured against.

Process Control and Verification

The measurements that matter are the repeatable ones: conductor width and spacing, annular ring, finished hole size, plating thickness and surface finish are all verifiable on a coupon that travels with the panel. Running a first article through the same checks as the production panel confirms that the two agree, and that comparison is the cheapest form of process control available at prototype stage.

Keeping a sample from the panel turns a dispute into a measurement, because the same coupon can be re-examined by both parties without rebuilding the batch. Reviewing the design before the data is released is cheaper than correcting it after the panel is in the tank, because every step downstream inherits the decision made at the front end.

Documenting the assumption is part of the design work, and a short note on the drawing prevents a question that would otherwise arrive a day later and cost a day of schedule. The process window is set by the narrowest step in the flow, so an improvement anywhere else shows up as margin rather than as yield until that step is addressed.

Is the slowest machine always the bottleneck? Not if it stops often or carries a heavy rework load. The effective cycle time is what decides.

Does a faster line mean a faster machine? It means less work at the constraint, which can come from quality, from setup or from a different machine.

What does gopcb provide for line capacity? We provide the constraint identified from cycle time, uptime and rework together, buffers sized from variability rather than from space, setup time analysed against batch size, and the product mix reflected in any stated capacity figure.

Checks Before Release

The checks that matter are the ones performed on the product rather than on a sample kept for the purpose, because a coupon that travels with the panel is the only evidence about that panel. Where a requirement can be measured, it should be measured at the point of manufacture and recorded against the board or the lot it applies to.

A parameter that is set once and never re verified drifts, and the drift is usually discovered by a defect rather than by the record. The tooling, the material and the profile form one system, and a change to any of them should be assessed against the other two before it is released.

Where the process window is narrow, the measurement resolution has to be better than the window, or the data cannot distinguish a good part from a marginal one. A record that identifies the operator, the date and the settings is worth more than a record that identifies only the result.