You deploy the control plan, train the operators, and pass the IATF 16949 surveillance audit. Six months later, you walk the floor and realise the documented process bears little resemblance to actual operations. Nobody decided to abandon the standard. The drift happened through a hundred invisible, unrecorded adjustments. This is quality entropy, and it actively dismantles quality management systems from the inside.
In physics, entropy is the measure of disorder in a closed system. The second law of thermodynamics dictates that systems naturally move from order to disorder unless energy is applied. Quality systems operate on the exact same principle. A perfectly engineered PFMEA and control plan will degrade unless the organisation continuously applies deliberate energy to maintain them.
I have audited plants that held impeccable documentation but produced nonconforming hardware because the shop floor had silently rewritten the process. Operators found faster methods, engineers forgot to update linked documents, and managers stopped questioning routine outputs. Entropy does not announce itself with a major defect. It compounds through minor, rationalised deviations until a failure occurs.
What Process Degradation Looks Like on the Floor
Quality entropy rarely shows up on your KPI dashboard until the damage is done. It operates through compounding micro-decisions. An inspector skips a non-value-added check on step seven because they have not found a defect in four thousand parts. That check becomes a glance, then a skip. When a defect finally slips through, the root cause is the absence of an action that nobody remembered stopping.
Consider the operator who discovers that skipping the in-process recording step saves twelve seconds per cycle. Over a month, they recover a full day of production capacity. The operator believes they are improving efficiency. The quality system assumes compliance. The reality is an undocumented process running outside the validated parameters, generating untraceable hardware that will fail during a customer audit.
This drift is not limited to direct labour. A daily quality huddle designed to review process parameters and corrective actions slowly degrades into a status reading. The phrase "everything looks good" stops being a data-driven conclusion and becomes a reflex. The meeting still happens on the schedule, but the critical thinking that gave the meeting value has evaporated.
The Primary Drivers of System Decay
The first driver is the normalisation of deviance. This occurs when organisations gradually accept abnormal conditions as standard because they have not yet caused a visible failure. You see this in unauthorised tolerance creep and informal reductions in inspection frequency. Temporary deviations become permanent fixtures, and new hires are trained on the drifted standard rather than the documented process.
The second driver is institutional forgetting. When a work instruction states "torque to 47 Nm ± 2 Nm," it dictates what to do, but not why. It omits that the tolerance was tightened after a field failure involving thread deformation. Without the engineering rationale behind the parameter, the specification looks like an arbitrary rule. Rules without context are the first targets for optimisation and elimination by well-meaning engineers.
The third driver is metric fatigue. When organisations track dozens of colour-coded parameters, the signals drown in noise. Operators and managers develop metric blindness. If a process runs with every indicator green while scrap costs rise, the metrics have lost their sensitivity and no longer reflect reality.
Documented Process vs Drifted Reality
Documented Process
- Record measurement data every cycle
- Inspect step seven for surface defects
- Tighten fastener to 47 Nm ± 2 Nm
- Review defect trends in daily huddle
Drifted Reality
- Record data only when a deviation occurs
- Visually glance at step seven occasionally
- Tighten until the gun clicks at roughly 45 Nm
- Read standard metrics without investigation
Change Management and Communication Breakdown
Every process exists inside a web of interconnected documents and dependencies. Uncontrolled change is a massive engine for entropy. An engineer updates a CAD model and revises the bill of materials, but the receiving inspection checklist references the old revision index. Six weeks later, the plant is inspecting parts against superseded dimensions, passing bad hardware, and rejecting good material.
Change without ripples occurs when modifications propagate unevenly across the management system. The engineering database is updated, but the PFMEA, the control plan, and the operator visual aid remain stagnant. Without a rigorous change management process that forces cross-functional updates, these disconnected documents create latent failures that auditors will eventually expose.
Communication decay accelerates this technical drift. Quality systems are fundamentally communication networks. When shift handovers become rushed and quality engineers stop walking the gemba, the immune system of the plant fails. The operators who used to flag minor concerns decide it is not worth the effort. When feedback loops break, every other failure goes undetected for much longer.
Measuring Drift: The Quality Degradation Index
You cannot fight what you cannot see. Establish a Quality Degradation Index (QDI) to periodically assess how closely actual operations match designed processes. The QDI is not a new audit standard. It is a diagnostic lens applied to existing IATF 16949 or AS9100 operations. It forces you to measure the gap between the paper system and the physical reality.
Process fidelity is the primary layer, carrying a thirty percent weight. Conduct unannounced process observations, not formal audits. Watch what operators actually do, not what they perform when they know they are being evaluated. Score critical processes on a scale of one to ten, where ten means the process is followed exactly as documented and one means the documentation is pure fiction.
Documentation currency and knowledge encoding form the next layers. Pull a random sample of active work instructions and control plans to verify they reflect current manufacturing reality. Then evaluate your top ten critical process parameters to see if the engineering rationale for their specific values actually exists. If the reasoning behind the parameters is not documented, the process is highly vulnerable to drift.
Quality Degradation Index Weighting
Building Anti-Entropy Practices
To fight back, implement process observation rotations. Once a month, have an engineer or manager spend thirty minutes observing an unfamiliar process. They are not there to audit or enforce compliance. Fresh eyes identify invisible workarounds that familiar operators take for granted. An innocent question about why a step is out of sequence often reveals a broken process that the team has normalised.
Pair this with a knowledge archive. For every critical process parameter, document the engineering story behind the specification. Record the exact failure mode that necessitated the tolerance, the cost of the associated nonconformance, and the technical reason the parameter was chosen. This narrative makes the specification immune to casual optimisation by engineers who do not remember the original field failure.
The moment you stop applying deliberate energy to a quality system, entropy begins dismantling it.
Implement scheduled process revalidation as a hard annual reset. Do not perform a document review. Run the most critical processes exactly as documented, measure the outputs, and compare the Cpk to the original PPAP validation. This forces the organisation to consciously evaluate whether materials, equipment, or environmental conditions have drifted from the validated baseline over the past twelve months.
For every engineering change, mandate change ripple mapping. Before closing a change order, verify that every linked document, visual aid, and inspection program has been updated. A simple cross-functional matrix prevents the all-too-common scenario where production launches a new revision while the quality lab inspects parts against a superseded drawing.
Leadership and the Discipline of Tension
Organisational systems take their cues from leadership behaviour. When plant managers accept "everything looks good" without questioning the underlying data, they actively endorse decay. When leaders skip the layered process audits or tolerate the gap between work instructions and reality because production targets are tight, they become entropy's greatest ally.
The most effective quality directors I have worked with share a common trait: they are perpetually dissatisfied with the gap between design and reality. They walk the floor, observe the actual flow, and ask operators to explain the parameters. They treat the discovery of a workaround not as a compliance failure, but as critical system information.
Here is the central paradox of quality management: the more successful your system is, the more invisible it becomes. When defects are rare, people forget why the prevention steps matter. When audits pass easily, people forget the rigour required to achieve those results. Quality is not a stable state you achieve and lock down. It is a tension you maintain against the constant pull of disorder.
Entropy cannot be defeated, but it can be managed through deliberate stewardship. Build observation, revalidation, and knowledge encoding into your operating rhythm. Maintain the tension, document the rationale, and never accept green dashboards without verifying the shop-floor reality. Your quality system is either actively resisting decay or quietly surrendering to it.