Pressure to produce more will always be part of manufacturing. New equipment, additional labor, and automation all have an important place in meeting that demand. The decision becomes much stronger when manufacturers first understand where the existing process is losing time
Lokesh Ramesh, Manufacturing Engineer
When a production line falls behind its target, the first response is often to look for more capacity. Can the equipment run faster? Can another operator be added? Would automation increase throughput? Does the plant need another machine?
Those questions may eventually be worth asking, but they come too early when a manufacturer does not yet understand where production time is being lost.
A line can appear to have a capacity problem when the real problem is a long changeover, uneven work distribution, excessive operator movement, recurring quality issues, material delays, or one process that consistently runs slower than everything around it. Increasing machine speed in that environment may simply produce more work in process without increasing finished output.
Follow the Time Through the Entire Line
Cycle time is a useful starting point, but a single average can conceal a great deal of variation.
Suppose a line is expected to produce one unit every 45 seconds. Several stations may consistently finish in 35 or 40 seconds while another requires 55. The faster stations are technically performing well, yet their additional speed contributes little if production continually queues at the slower operation. This is where line balancing becomes critical. The goal is to understand how work is distributed across the process and whether each station can support the required production rate.
The same analysis should extend beyond machine cycle time. Operator movement matters. An operator who repeatedly walks across a work area to retrieve components, tools, labels, or packaging is consuming production time even when the equipment itself is running perfectly. Waiting for material replenishment, inspections, approvals, or upstream processes can have the same effect.
Lean manufacturing has long treated unnecessary motion and waiting as sources of waste. The National Institute of Standards and Technology has examined inventory and flow time in United States manufacturing, including the ways production systems accumulate time between activities rather than only during the manufacturing operation itself.
On one solar module line I took over, the stringing machines showed more recorded downtime than the next four areas combined, so they were assumed to be the problem. Through Gemba walks and cycle time analysis, I found that the stringers were actually being stopped because downstream equipment could not move modules fast enough.
The stringers could produce 240 modules per hour, while the slowest downstream station, the auto turn, had an 18 second cycle time that limited theoretical capacity to about 200 modules per hour. The conveyor and encapsulant sections were similarly constrained, while equipment farther downstream was comfortably faster.
I synchronized material handling transfer speeds with downstream machine cycle elements across the auto turn and encapsulant sections, reducing the bottleneck cycle time from 18 seconds to approximately 17 seconds and increasing line capacity to roughly 215 modules per hour. Over the following ten days, recorded downtime fell 12 percent and runtime increased 13 percent.
The purpose of the exercise is to understand which activities are controlling overall output.
Separate the Symptom From the Constraint
Production data becomes much more useful when it helps explain why a target was missed.
Overall equipment effectiveness, commonly called OEE, provides one useful view because it considers availability, performance, and quality together. Vorne explains OEE as a framework for measuring productive manufacturing time through availability, performance, and quality.
Looking at those components individually can reveal very different problems.
A performance loss may indicate that equipment is running below its intended rate. An availability loss may point toward downtime, adjustments, maintenance, or changeovers. A quality loss may mean the line is producing at an acceptable rate while losing capacity through scrap and rework.
First pass quality is particularly important when evaluating throughput. A unit that reaches the end of a process and then requires additional work has consumed production capacity without producing a finished product. If that same unit must reenter the process, it consumes capacity again.
Increasing upstream speed can make this problem worse because the line may simply generate defective or incomplete work more quickly.
Structured continuous improvement methods are useful because they force teams to investigate the process before selecting the solution. One of the most established is DMAIC, which stands for define, measure, analyze, improve, and control. ASQ describes DMAIC as a structured approach for improving existing processes that are not meeting required performance levels.
The measure and analyze phases are especially important in manufacturing environments. Teams can establish baseline performance, map the current process, examine variation in cycle times, identify recurring downtime, study quality losses, and determine which factors have the greatest influence on output.
That work prevents teams from confusing a visible symptom with the actual constraint.
For example, a line that repeatedly misses its daily production target may appear to have an equipment speed problem. Data may instead show that the largest losses come from startup delays, frequent minor stops, material handling, or a quality issue concentrated at one operation.
On one production line, final inspection yield was barely above 96 percent and modules were repeatedly piling up at repair. The assumption was that the repair station needed more people, but the real problem was repeat defects. Modules that left repair still defective cycled back through inspection, filled the buffer, stopped the conveyors, and ultimately slowed the stringers. In OEE terms, the line was treating a quality loss like a performance problem.
During the measure phase, I began tracking defects by operator rather than only by station and traced each downgraded module back through its inspection image to the person who had repaired it. That revealed variation that station level averages had been hiding and showed that handling technique, rather than machine capability, was driving the loss.
Targeted retraining reduced final inspection defects from repaired modules from 182 to 76 modules within three weeks, a 58 percent reduction. Operator induced gap defects fell by roughly 40 percent, soldering defects dropped from more than 500 modules in May to fewer than 10 modules in September, and final inspection yield improved from about 96 percent to just over 99.2 percent over the year. No equipment was purchased and no additional headcount was added. The improvement came from better measurement and targeted training.
Once a team can identify exactly where production time is being lost, improvement becomes far more focused.
Recover Existing Capacity Before Buying New Capacity
Automation can create significant value in manufacturing, particularly when work is repetitive, physically demanding, inconsistent, difficult to scale, or dependent on information that can be processed more reliably by a system.
The existing workflow still determines how much value that automation can create.
Automating an inefficient sequence preserves the sequence. Adding equipment upstream from a bottleneck can increase queues. Adding labor without understanding workload distribution can place people where additional capacity has little effect. Running a machine faster can create pressure farther down the line that eventually appears as downtime, excess inventory, quality loss, or waiting.
This is why process improvement should come before major capacity decisions whenever possible.
Before investing in additional equipment, labor, or automation, manufacturers should be able to answer several practical questions:
- Which operation currently constrains total output?
- How much production time is lost to changeovers and planned stops?
- Where do operators spend time walking, waiting, searching, or handling material?
- How much capacity is consumed by defects and rework?
- Are cycle times balanced closely enough to support the required production rate?
- Which losses occur frequently enough to justify improvement?
- What measurable production outcome should change when the problem is solved?
- A manufacturer that understands its constraints can automate with a specific objective and then measure whether the investment actually produced the expected improvement.
The Manufacturing Extension Partnership at NIST emphasizes continuous improvement as an important part of improving productivity and competitiveness. That principle becomes especially valuable when manufacturers are under pressure to increase output while controlling capital and labor costs.
On the stringer line, an automated labelling station became the next front end constraint. The full cycle took about 12.5 to 13 seconds, followed by another 5 to 5.5 seconds of robot idle time while the next module arrived and was scanned.
The issue was the module barcode scanner location. Because it is located at the end of the labelling conveyor, the next label could not be printed until the module was already in position. Moving the scanner to the preceding conveyor allowed the label to be prepared before arrival, removing roughly 5 seconds of idle time from each cycle, or close to 30 percent of the station time, without adding equipment or labour.
The change was validated and approved for implementation, creating a clear path to higher throughput and more consistent production.
Those outcomes tell a plant much more than equipment speed alone.
Pressure to produce more will always be part of manufacturing. New equipment, additional labor, and automation all have an important place in meeting that demand. The decision becomes much stronger when manufacturers first understand where the existing process is losing time.
Once those losses are visible, manufacturers can decide whether they truly need more capacity or whether the first opportunity is to recover the capacity they already have.
Lokesh Ramesh is a Manufacturing Engineer, and a Full Member of Sigma Xi, The Scientific Research Honor Society. His work focuses on Lean Six Sigma, production optimization, advanced materials, and industrial quality systems, with experience improving OEE and reducing costs across high volume manufacturing environments. He holds an MS in Materials Science and Engineering from the University of California, Riverside, and is the inventor on two U.S. patent-pending technologies.
