Know Your Manufacturing Process Baseline Before Improving

Before diving into advanced line upgrades or expensive automation, the first step is to understand the manufacturing process baseline. What do you change if you do not know what needs to be changed and why? This involves collecting detailed data on how a manufacturing line currently performs — from cycle time to unwanted variation from the specific process — and evaluating its current capabilities and limitations. Without this foundational step, improvement efforts risk being inefficient, misaligned, or even counterproductive. I have been thinking about this because of flashbacks to some work I did over the course of my career, as well as recent consultations. In fact, this chat about measurements and controlled and discipline approach to exploration applies to product testing as well.

Why the Manufacturing Process Baseline Matters

To begin with, if you are familiar with APQP (Advanced Product Quality Planning), you know about PFMEA (Process Failure Mode and Effects Analysis) and the connection to the Control Plan. A Process Failure Mode and Effects Analysis (PFMEA) is a structured approach to identifying and prioritizing potential failure modes within a manufacturing process. It evaluates the severity, occurrence, and detection of each failure to calculate a Risk Priority Number (RPN), guiding teams to focus on the most critical risks. The insights from the PFMEA directly drive the development of the Control Plan, which outlines how the process will be monitored and controlled to prevent or mitigate those risks. The Control Plan operates across three levels: prototype, pre-launch, and production. The prototype control plan focuses on early design validation and high-level risk identification. The pre-launch control plan adds more detail, incorporating data from trial builds and refining controls based on actual performance. The production control plan is the most comprehensive, detailing all monitoring systems, inspection methods, reaction plans, and feedback loops needed to ensure ongoing process stability and quality. Together, PFMEA and the three-tiered Control Plan create a systematic path from risk identification to robust process control.

Essentially, we learn how to build the product throughout the product development effort. We make decisions about the manufacturing line based on our critique of the proposed manufacturing line, as well as parts from the manufacturing line along the way, via prototype builds, trial production runs, run-at-rates, and PVT (Process Validation Testing). Or at least we should be learning. What happens when the manufacturing line has been running for many years, and new people join?

Knowing the manufacturing process baseline gives a clear, data-driven snapshot of current operational effectiveness. It highlights:

Critical to Quality (QTC) attributes
Process variability
Special tools (gauges)
- Key measurements
- Gauge R and R (variation of, product, measuring instrument, and interpretation of the measurement)
- Calibration and maintenance
Bottlenecks
Defect frequency and type
First Pass Yield (FPY) percentages
Capability indices (Cp, Cpk)

These metrics are essential for any meaningful process improvement, ensuring that updates align with actual production realities rather than assumptions.

Connecting Baseline to TQM and PDCA

Total Quality Management (TQM) is built on data-informed continuous improvement. The Plan-Do-Check-Act (PDCA) cycle supports that by offering a structured approach:

Plan: Collect baseline data. Define problems and plan experiments.
Do: Implement small-scale improvements based on the process baseline.
Check: Analyze outcomes using the original baseline for comparison. Are results stable or shifting? Were our predictions accurate?
Act: If our predictions are accurate, we want to implement effective changes; if results deviate, either for the better or worse, we need more analysis as to why and perhaps return to planning.

The “Check” step is crucial — this is where predictability is tested. If results from TQM initiatives vary widely or outside of our predictions, we have more learning to do. I often write data without context, and iteration offers little value. I emphasize the need for measurement before modification.

Predictability and Capability: Are Results Stable?

Using statistical process control (SPC), process behavior can be charted before and after changes. If outcomes remain within expected control limits and show less variability, the process is predictable. If not, root causes need further analysis. If our predictions are not accurate, even if performance is better than we anticipated, we need to spend some time figuring out why. The basis for our improvement initiative is shaky. At the very minimum, we need to revisit the available data to see what we missed.

This level of scrutiny ensures each improvement is genuinely effective and not a fluke of random variation. Our writings stress that change should lead to measurable, repeatable gains, not just perceived improvement.

Improving First Pass Yield Starts with the Baseline

First Pass Yield (FPY) is a primary metric for process efficiency. Without an accurate manufacturing process baseline, FPY goals are arbitrary. Improvement initiatives must be grounded in how the line performs today.

Once the baseline is defined, lean tools — 5 Whys, Ishikawa diagrams, and process mapping — can identify the root causes of low FPY. Only then can potential sustainable improvements be explored, verified, and then introduced into production.

Summary: Don’t Upgrade Blind — Analyze First

Improving capability and FPY starts not with upgrades, but with understanding. Establish the manufacturing process baseline, apply the PDCA cycle, and follow TQM principles. As I highlight, process knowledge is the foundation of innovation. Our team learning along the way is what makes this work, not the buzzwords TQM or PDCA. We need to understand what we have, and that happens through exploration AND DATA!