In thermodynamics, the second law dictates that closed systems naturally drift towards disorder. Quality management systems obey a similar, unavoidable rule. The documentation you wrote three years ago no longer matches actual shop-floor execution. The control limits calculated during your initial PPAP submission have drifted, yet nobody has recalculated them.
The training delivered during line launch is increasingly irrelevant. Half the original operators have left, and their replacements learned the process from people who were already doing it incorrectly. This is quality entropy. If you do not build active mechanisms to fight it, you will spend your career reacting to a force you cannot initially see.
Quality entropy is the gradual, invisible degradation of your standards, processes, and controls. It happens because there are vastly more ways to execute a process incorrectly than correctly. Maintaining a validated state is not a passive exercise. It requires a constant, measurable input of energy to hold disorder at bay.
The Mechanics of System Degradation
Organisations pay an entropy tax whether they acknowledge it or not. The question is never whether a system degrades, but how you choose to pay the tax. You either pay in small, controlled instalments through active maintenance, or you pay a catastrophic lump sum during a recall, an IATF 16949 customer audit failure, or an EASA regulatory shutdown.
I have audited plants that passed their initial certification flawlessly, only to find critical failures eighteen months later. Control charts were still being filled out manually, but the statistical control limits had not been recalculated for over a year. Incoming inspection protocols were ignored entirely because a critical supplier had quietly changed their manufacturing process.
No single person decided to sabotage these systems. No executive mandated the budget cuts. The quality system simply drifted. Like a vessel with no one at the helm, it moved away from its validated course through a thousand imperceptible shifts. Each individual shift was small enough to ignore during the daily production rush. Together, they created systemic risk.
Knowledge Decay and Process Mutation
Every organization loses tacit knowledge faster than it acquires it. The procedure documents on your server capture perhaps thirty percent of what experienced operators actually know. When senior operators retire or transfer, that unrecorded seventy percent vanishes from the living memory of the floor.
I recall auditing an automotive supplier where a critical heat treatment process ran perfectly for five years. The original engineer had retired. When a subtle change in furnace behaviour occurred, nobody recognized the signal. The process drifted for three weeks before the customer caught the nonconformance during incoming inspection.
Processes also mutate without formal change requests. A supervisor allows a temporary deviation that becomes a permanent, unrecorded standard. A maintenance technician adjusts a pneumatic pressure setting and forgets to log it. Organizational environments naturally select for speed and convenience over strict precision and compliance.
Unlike biological evolution, where harmful mutations are eliminated, factory floor mutations survive precisely because they make the operator's life easier in the short term. These undocumented shortcuts bypass the PFMEA risk controls entirely. They accumulate until a severe defect escapes to the customer.
Measurement Erosion and Attention Dissipation
The measurement systems you rely on to detect degradation are themselves subject to degradation. Gauges drift physically. MSA studies become invalid when new, untrained inspectors are assigned. A go/no-go gauge that was perfect when purchased may have been damaged repeatedly, rendering its accepted parts mathematically unreliable.
Meanwhile, the metrics chosen to monitor your system were selected at a specific moment for a specific set of conditions. As volumes scale and product mixes change, those metrics become irrelevant. You continue measuring OEE and scrap rates because they are what you have always measured, entirely missing the drift occurring in dimensions you never thought to track.
Quality requires sustained attention, which is a strictly finite resource. When a new priority emerges, leadership attention shifts. The 8D corrective action reviews that were rigorous six months ago suddenly receive only cursory oversight. Nobody decided quality was less important; the daily crisis simply became louder than the established routine.
Quality maintenance is not the cost of doing quality; it is the strategy for sustaining it against constant degradation.
The Entropy Tax Payment Models
The Lump-Sum Approach
- Validation treated as a one-time launch event
- Attention diverted to immediate production fires
- Process drift goes unmeasured and unnoticed
- Result: Recall, audit failure, or regulatory action
The Active Maintenance Approach
- Scheduled revalidation based on risk severity
- Dedicated resources for layered process audits
- Control limits recalculated on a fixed cadence
- Result: Predictable costs and stable Cpk performance
Scheduled Revalidation and Knowledge Redundancy
Most organizations validate a process once during initial launch and assume it remains validated forever. This is equivalent to assuming a machine remains permanently aligned because it left the factory floor within tolerance. You must build a revalidation schedule based on actual risk, not arbitrary convenience.
High-risk processes—such as welding, heat treatment, or sterile packaging—require annual revalidation. Medium-risk processes require a biennial check. Revalidation does not need to match the scope of the original PQ, but it must verify that the process still meets the exact same acceptance criteria established during the initial run.
Simultaneously, you must force knowledge redundancy. For every special process, ensure that explicit knowledge is distributed across at least three people who actively maintain it. Do not count three people who once signed a training record. Count three people who could recreate the process setup from scratch if the documentation system crashed.
Use structured knowledge transfer sessions to achieve this. Have the subject matter expert run the line while a novice documents the step-by-step procedure. Compare the novice's fresh documentation to the existing control plan. The gaps between them reveal exactly where knowledge entropy has already occurred.
Making Degradation Visible
Entropy is most dangerous when it operates invisibly. You must build early-warning systems that make degradation visible long before it becomes a critical defect. Statistical process control is one such system, but only when control limits are recalculated on a defined schedule, not left static for years.
Layered process audits provide another data channel. You cannot see drift if you do not remember your baseline. Archive your initial process capability studies and compare current performance to those historical Cpk targets. Keep the original control plans accessible and use them as physical references during floor audits.
Visibility requires deliberate comparison. Maintain visual examples of what acceptable output looks like. When an inspector or operator reviews a part, they need an immediate, physical reference point to compare it against, rather than relying on memory or a dim awareness of a specification they read months ago.
Anti-Degradation Cadence Targets
Maintenance as a Strategic Imperative
The organizations that sustain quality excellence over decades do not treat system maintenance as an administrative chore. They treat it as a core strategic function. They allocate specific budgets for calibration, training, and internal audits. They assign their most competent engineers to manage these programmes.
Leaders set the energy budget for the organization. When a plant manager signals that management reviews, training days, or cross-functional audits are less critical than hitting a shipment target, they are actively reducing the energy available to fight entropy. The system will degrade. The only variable is the speed of the collapse.
The best quality leaders treat audits as system health checks, not as bureaucratic interruptions. They view training as insurance against knowledge decay. They demand unglamorous, disciplined maintenance work even when nothing appears to be wrong, because they understand that invisibility is exactly what makes entropy dangerous.
Quality entropy is an observation confirmed by every corrective action and every failed surveillance audit in the history of manufacturing. The only defense is constant, intentional energy. The standard you set today is already degrading. You must build the systems to measure the drift before you are surprised by the collapse.