Predictive Maintenance vs Reactive Maintenance: What Nigerian Manufacturers Need to Know

Taiwo manages the operations of a medium-sized flour milling company in Ibadan. His facility runs three roller mills, two plansifters, and an assortment of conveying, bagging, and dust extraction equipment that together form a continuous production process from raw wheat intake to finished packaged flour. The facility runs two twelve-hour shifts, six days a week. It cannot afford to stop unexpectedly, because flour buyers in the retail and commercial bakery market operate on just-in-time restocking, and a supplier who cannot deliver on schedule loses shelf space and bakery contracts to competitors who can.

Over a twelve-month period that Taiwo calculated with some pain in retrospect, unplanned equipment stoppages cost his facility eleven production days in total across seven separate breakdown incidents. The incidents ranged in severity from a three-hour conveyor belt failure to a two-and-a-half-day roller mill shutdown caused by a gearbox failure that required a replacement part to be sourced from a specialist supplier in Lagos. The combined cost of lost production, emergency repairs, expedited parts sourcing, and one customer penalty clause came to just over two million, eight hundred thousand naira. When Taiwo's plant engineer showed him this figure, assembled for the first time as a single annual breakdown cost, his first question was not how to reduce repair costs. It was: could any of these breakdowns have been predicted and prevented?

The answer, his engineer told him, was that at least five of the seven incidents involved equipment that had been showing detectable signs of deterioration for weeks before it failed. The gearbox that eventually seized had been running at elevated temperature for at least three weeks prior to failure, a condition that would have been visible with a handheld infrared thermometer during any routine inspection. The conveyor bearing that failed had been producing an intermittent high-pitched sound that several operators had noticed and mentioned informally to each other but had not formally reported. Two of the remaining incidents involved belts and seals that had been visibly worn at the last informal maintenance inspection but had not been replaced because there was no spare in stock and no formal process for actioning the observation.

Taiwo's experience sits precisely at the intersection of the two maintenance philosophies this article addresses. His facility had been operating on reactive maintenance: waiting for equipment to fail before attending to it. His engineer was describing what predictive maintenance would have given him: the ability to detect developing failures in advance and act before they become production stoppages. Understanding the real difference between these two approaches, what each one actually requires, what it costs, and what it delivers in the specific context of Nigerian manufacturing, is what this article is designed to provide.

Reactive Maintenance: An Honest Assessment

What Reactive Maintenance Actually Looks Like in Practice

Reactive maintenance is the default maintenance philosophy of most Nigerian manufacturing facilities, and it is worth being precise about what it actually entails before criticising it, because its persistence is not irrational. In a reactive maintenance model, equipment is operated until it fails or until a fault becomes obvious enough to trigger an unscheduled stoppage. The maintenance team then responds to the failure, diagnoses the cause, sources the required parts or skills, and restores the equipment to operation. The focus of the maintenance function is almost entirely on repair speed: how quickly the broken machine can be returned to production.

This approach has genuine advantages that explain why it remains so widespread. It requires no upfront investment in inspection equipment, diagnostic technology, or analytical capability. It demands no maintenance planning effort during periods when equipment is running smoothly. It produces immediately visible outcomes, because every repair is a tangible, measurable event, whereas prevented breakdowns are invisible by definition. For equipment that is genuinely low-criticality, inexpensive to repair, and fast to restore to service, reactive maintenance is entirely appropriate and economically rational. Nobody seriously argues that every light bulb in a factory should be on a preventive replacement schedule. Reactive replacement of non-critical, low-cost, easily substituted items is a perfectly sensible maintenance strategy.

The problem is not reactive maintenance as a philosophy. The problem is applying reactive maintenance to equipment that does not fit those three conditions: equipment that is high-criticality, expensive or slow to repair, and not easily substituted while it is down. When reactive maintenance is the strategy for this category of equipment, the results are the ones Taiwo experienced: unpredictable, severe, and expensive stoppages that arrive without warning and produce the maximum possible operational and commercial damage before they are resolved.

The Hidden Costs That Make Reactive Maintenance Expensive

The financial case against reactive maintenance for critical equipment is not simply that repairs are expensive. It is that reactive repair is consistently more expensive than the equivalent preventive intervention would have been, for reasons that are structural rather than accidental. When a component fails catastrophically, it frequently damages adjacent components in the process of failing. A bearing that seizes does not simply stop working. The heat and mechanical stress of its seizure can damage the shaft it sits on, distort the housing it is mounted in, and in some cases damage gears, seals, or alignment fixtures in the surrounding assembly. The repair bill for a reactive bearing failure is therefore not just the cost of a replacement bearing. It is the cost of the bearing plus the cost of repairing everything the failed bearing damaged on its way out, which is often several times more expensive than the bearing itself.

Beyond the expanded repair scope, reactive maintenance carries the costs of emergency economics. When a critical production machine fails unexpectedly, the normal procurement process for spare parts is bypassed in favour of whoever can supply the needed component fastest, regardless of price. Emergency suppliers command premium pricing. Expedited delivery from Lagos or from abroad carries costs that planned procurement does not. Specialist technicians called in urgently rather than scheduled in advance charge higher rates and may incur travel and accommodation expenses. Every element of a reactive repair costs more than the equivalent element of a planned repair, and in Nigerian manufacturing, where spare parts are not universally locally available and where specialist expertise is geographically concentrated, these emergency premiums are particularly steep.

The third hidden cost is the opportunity cost of the production time lost during an unplanned stoppage. This is the cost that most reactive maintenance accounting ignores but that is typically the largest single component of the total breakdown cost. Every hour a critical production line stands idle, the factory's fixed overhead continues to accumulate: staff wages, generator fuel, facility rent, loan repayments, and management salaries all tick on regardless of whether product is flowing out of the factory. Against this running cost, zero revenue is being generated, because a stopped line produces nothing to sell. In a factory with a high fixed cost structure, as most Nigerian manufacturing businesses are, the overhead absorption cost of a single day of unplanned downtime can dwarf the actual repair bill for the failure that caused it.

When Reactive Maintenance Is the Right Choice

Before leaving reactive maintenance behind entirely, it is important to be explicit about the circumstances in which it remains the correct strategy. Not every piece of equipment in a Nigerian factory warrants preventive or predictive maintenance attention, and attempting to apply structured maintenance programmes to every asset regardless of criticality wastes maintenance resources that should be concentrated where they deliver the greatest return.

The appropriate candidates for reactive maintenance are equipment items whose failure has no significant impact on production continuity, either because they are non-critical to the main production process or because an immediate substitute is available. They are items whose repair cost is low and whose repair time is short, making the operational disruption of a reactive response negligible. They are items whose failure mode is genuinely random, meaning that no amount of inspection or monitoring would provide meaningful advance warning, making preventive or predictive intervention wasteful rather than beneficial. For these items, running to failure is not a maintenance failure. It is a rational economic decision, and the maintenance manager who applies the same intensive monitoring to a packaging room fan as to a primary production drive motor is misallocating scarce maintenance capacity.

The practical implication of this distinction is that every Nigerian manufacturing facility should have an explicit categorisation of its equipment by maintenance strategy, rather than applying a single approach uniformly across all assets. Critical production equipment warrants preventive or predictive maintenance. Moderately important equipment warrants preventive maintenance at appropriate intervals. Non-critical equipment can appropriately be maintained reactively. Building this equipment strategy map is one of the most valuable exercises a factory maintenance manager can undertake, because it aligns maintenance effort with maintenance value in a way that neither pure reactive nor uniformly preventive approaches can achieve.

Predictive Maintenance: What It Really Means

The Fundamental Idea

Predictive maintenance is an approach to equipment management that is based on a single foundational insight: most equipment failures do not happen instantaneously. They develop over time, preceded by a measurable deterioration in the equipment's condition that, if detected early enough, provides an opportunity to intervene before the deterioration reaches the failure threshold. A bearing does not simply fail. It develops increasing internal wear that produces rising vibration and temperature. A motor winding does not simply burn out. Its insulation resistance gradually degrades in ways that are measurable with appropriate diagnostic equipment. A gearbox does not simply seize. Wear particles from its gear surfaces accumulate in its oil in increasing concentrations that oil sample analysis can detect weeks or months before the gearbox fails.

Predictive maintenance uses condition monitoring technologies to measure these early indicators of deterioration on a regular basis, building a picture of each equipment item's health over time and identifying the point at which a measured parameter has moved far enough from its baseline to indicate that maintenance intervention will be needed before the next scheduled inspection. The maintenance action is then planned and executed at a time that is convenient for production, with the correct parts and skills prepared in advance, before the deterioration reaches the failure point. The result is that maintenance activity is driven by the actual condition of each piece of equipment rather than by a fixed calendar schedule or by the occurrence of a failure.

The Main Condition Monitoring Technologies

Vibration analysis is the most widely used condition monitoring technology for rotating machinery, which encompasses the large majority of industrial production equipment: motors, pumps, fans, compressors, gearboxes, and all rolling element bearings. Every rotating machine produces vibration as it operates, and the frequency and amplitude patterns of that vibration change in characteristic ways as different types of mechanical fault develop. A bearing with developing internal wear produces vibration at specific frequencies determined by the geometry of the bearing and its rotational speed. A gear with a developing tooth defect produces vibration at the gear mesh frequency and its harmonics. Vibration analysis equipment, ranging from simple handheld vibration meters that cost a fraction of the price of a single major repair to sophisticated multi-channel online monitoring systems, captures these vibration signatures and allows trained analysts to identify developing faults and estimate how much time remains before intervention is needed.

Thermal imaging uses infrared cameras to measure the surface temperature distribution of equipment without physical contact. Electrical equipment under stress, mechanical equipment with inadequate lubrication or bearing damage, and heat exchangers with fouled surfaces all produce characteristic thermal signatures that are invisible to the naked eye but clearly visible in a thermal image. A thermal scan of a switchgear panel will reveal loose connections and overloaded conductors as hot spots long before they produce an electrical fault. A thermal scan of a motor drive end will reveal a bearing running hot from inadequate lubrication or developing internal wear. Thermal imaging cameras have become considerably more affordable over the past decade, and periodic thermal surveys of electrical panels, motor housings, and mechanical drive systems are now a cost-effective condition monitoring tool for Nigerian manufacturers at almost any scale of operation.

Oil analysis involves taking small samples of the lubricating oil from gearboxes, hydraulic systems, and large rotating machinery and sending them to a laboratory for analysis of their chemical and physical properties. The concentration of metallic wear particles in the oil, the viscosity and acidity of the oil itself, and the presence of contaminants such as water or fuel all provide information about the condition of the machine internals that no external inspection can reveal. A gearbox whose oil sample shows rapidly increasing iron and chromium particle concentrations is a gearbox whose gear surfaces are wearing faster than normal, a condition that may be several months away from producing a failure but that warrants investigation and planned maintenance now rather than a reactive repair later. Oil analysis is particularly valuable for enclosed rotating machinery where direct visual or vibration inspection cannot easily access the critical wear surfaces.

Ultrasonic testing uses high-frequency sound waves to detect developing faults in a range of equipment types. Compressed air and steam leaks, which are significant energy waste sources in many Nigerian factories, are detectable with ultrasonic instruments at distances that make systematic survey of an entire facility practical. Early-stage bearing damage produces ultrasonic frequencies that appear before the vibration signatures detectable by standard vibration instruments, making ultrasonic monitoring a useful early warning tool for bearing condition. Electrical discharge in high-voltage equipment, which indicates insulation breakdown that precedes electrical failure, is detectable ultrasonically from a safe distance. For Nigerian manufacturers with significant compressed air or steam infrastructure, ultrasonic leak detection alone often pays for the equipment through the energy savings identified in the first survey.

What Predictive Maintenance Requires to Work

Predictive maintenance is not simply a matter of acquiring condition monitoring equipment. The technology is the tool, but the programme that makes the technology valuable requires several additional elements that are less obviously visible but equally important. The first is a baseline dataset: measurements taken on each monitored asset when it is in known good condition, which provide the reference against which future measurements are compared to determine whether a condition has changed. Without a baseline, a single vibration or temperature reading tells you very little. Compared against the baseline for that specific asset at that specific operating condition, it tells you whether the machine is getting worse, staying the same, or, occasionally, improving after a maintenance intervention.

The second requirement is measurement consistency. Vibration readings taken at different points on a machine, at different operating loads, or with different measurement techniques cannot be meaningfully compared over time. A predictive maintenance programme requires standardised measurement routes, defined measurement points on each machine, specified operating conditions under which measurements are taken, and consistent technique from the personnel conducting the measurements. This consistency requirement means that whoever conducts the condition monitoring surveys must be trained not just in the use of the equipment but in the importance of following a standardised procedure every time, even when production pressures make shortcuts tempting.

The third requirement is the analytical capability to interpret what the measurements are saying. Raw vibration data, a thermal image, or an oil analysis report is not a maintenance decision. It is information that requires interpretation by someone who understands what normal looks like for that equipment type, what the characteristic signatures of different fault types look like, and how to assess the severity and urgency of any anomaly detected. This analytical expertise can be built internally through training and accumulated experience, or it can be sourced externally from specialist condition monitoring service providers who conduct periodic surveys and provide interpreted reports. For smaller Nigerian manufacturers who cannot justify a full-time condition monitoring analyst, the external service provider model is a practical way to access predictive maintenance capability without building the full internal skill set from scratch.

Comparing the Two Approaches: What the Numbers Say

The Cost Structure Difference

The financial comparison between predictive and reactive maintenance is not simply a comparison of repair costs. It is a comparison of two fundamentally different cost structures, each with its own profile of fixed investments, variable expenses, and financial risks. Reactive maintenance has a low fixed cost structure: it requires no upfront investment in monitoring equipment, no training budget for analytical skills, and no planned maintenance labour cost during periods when equipment is running smoothly. Its variable costs are high and unpredictable: repair costs when failures occur, emergency premium costs associated with unplanned sourcing and labour, and production downtime costs that fluctuate with the severity and frequency of breakdown incidents. The reactive maintenance budget is hard to predict because its largest cost components are driven by events that happen on an unplanned schedule.

Predictive maintenance has a higher fixed cost structure: it requires investment in condition monitoring equipment, training for the personnel who conduct surveys and interpret results, and a planned maintenance labour allocation for the monitoring programme itself. Its variable costs are lower and more predictable: planned repairs are executed with correct parts and skills prepared in advance, at standard rates, during scheduled production windows, without the emergency premiums that reactive repairs carry. The predictive maintenance budget is more predictable because the majority of its maintenance interventions are planned rather than reactive, and planning enables competitive sourcing, appropriate skill mobilisation, and production schedule coordination.

For critical equipment in a high-throughput Nigerian factory, the economics of this comparison are strongly in favour of predictive maintenance once the fixed cost investment has been made. The reduction in emergency repair premiums, the elimination or near-elimination of the catastrophic multi-day breakdown events that carry the highest total cost, and the extension of component life through timely planned replacement rather than run-to-failure destruction, typically deliver financial returns that exceed the cost of the monitoring programme by a factor of three to five in the first full year of operation. These returns compound over time as the programme develops richer baseline data, more accurate deterioration trend analysis, and a track record of avoided breakdowns that becomes increasingly compelling to management.

The Risk Profile Difference

Beyond the direct financial comparison, reactive and predictive maintenance carry different risk profiles that matter enormously in the Nigerian manufacturing context. Reactive maintenance carries high tail risk: the small probability but very real possibility of a catastrophic failure that produces not just lost production but physical damage to the factory, injury to workers, or destruction of a machine that takes months to replace because a replacement is not locally available. These catastrophic outcomes are rare, but their probability is not zero, and in a country where insurance coverage for business interruption is often inadequate and where the financial reserves to absorb a major unexpected capital replacement are limited, the tail risk of reactive maintenance is a genuinely serious concern for the long-term financial stability of the business.

Predictive maintenance substantially reduces this tail risk because it specifically targets the failure modes most likely to produce catastrophic outcomes, the large rotating machines and high-energy systems whose failures carry the greatest potential for secondary damage and extended downtime, and keeps them under continuous condition surveillance. A factory that has been monitoring the vibration signature of its primary drive gearbox for twelve months has almost no chance of being surprised by a sudden catastrophic gearbox failure, because the deteriorating vibration trend that precedes such a failure will have been visible and actionable for weeks or months before it reaches the failure point. The risk profile of a predictive maintenance programme is one of many small, planned, manageable maintenance interventions rather than occasional large, unplanned, potentially catastrophic failures.

The Equipment Life Difference

One financial benefit of predictive over reactive maintenance that is consistently undervalued in short-term cost comparisons is the extension of equipment service life that condition-based maintenance delivers. Machinery that is maintained reactively is typically run until its components fail, at which point the failure event itself may have caused secondary damage that reduces the remaining serviceable life of the machine. Over multiple breakdown and repair cycles, cumulative secondary damage accumulates in the machine's structure, reducing its reliability, its performance, and eventually its useful life below what it could have been with more attentive maintenance.

Machinery maintained predictively, with components replaced based on measured condition rather than calendar schedule or failure occurrence, experiences far less secondary damage and far fewer repair cycles of the destructive reactive variety. Each planned replacement of a deteriorating component before it fails is a preservation of the machine's structural integrity. The gearbox whose oil is changed on the basis of oil analysis showing degraded additive chemistry and rising wear particles, rather than on a fixed calendar interval, receives exactly the maintenance it needs at exactly the time it needs it, and is not subjected to either the damage of continued operation on degraded lubricant or the unnecessary cost of an oil change on a machine whose oil is still perfectly serviceable. This precision maintenance extends the productive life of expensive industrial equipment in ways that accumulate significant capital value over the lifetime of a manufacturing facility.

The Nigerian Manufacturing Context: Special Considerations

The Operating Environment and Its Effect on Deterioration Rates

Predictive maintenance programmes developed in temperate manufacturing economies are typically calibrated to the deterioration rates and failure modes that equipment experiences under those conditions. Nigerian factories operate under conditions that accelerate many deterioration processes, which has practical implications for how a predictive maintenance programme should be configured to be effective in the local context. High ambient temperatures increase the operating temperature of motors, gearboxes, and control panels above the levels experienced in cooler climates, accelerating insulation degradation, lubricant oxidation, and bearing wear. Dusty environments, which characterise many Nigerian industrial locations, increase the contamination load on air filters, cooling systems, and lubricants, and also create abrasive wear mechanisms in machines that are not adequately sealed against dust ingress.

The practical implication is that measurement intervals for condition monitoring in Nigerian factories should generally be shorter than the equivalent intervals recommended in international standards or OEM guidance, and alarm thresholds for condition changes should be reviewed in light of the faster deterioration rates that local conditions can produce. A bearing that would be monitored monthly in a temperate European factory might warrant fortnightly vibration checks in a hot, dusty Nigerian production environment. A motor winding insulation resistance test that would be conducted annually as part of a standard preventive maintenance programme in a mild climate might benefit from quarterly measurement in a facility where high ambient temperature and humidity accelerate insulation ageing. Building these local calibrations into the programme design from the outset produces a system that is genuinely fit for the environment in which it operates.

The Spare Parts Supply Chain and Predictive Maintenance

One of the most practically important advantages of predictive maintenance in the Nigerian context is the transformation it produces in spare parts management. In a reactive maintenance world, spare parts are purchased in response to failures, which means they are purchased urgently, from whoever can supply them fastest, at emergency pricing, with lead times that determine how long the machine stands idle. In a predictive maintenance world, spare parts are purchased in response to condition monitoring data that indicates a component will need replacement within a defined time horizon, typically measured in weeks or months. This planned horizon creates the opportunity to source parts through normal procurement channels, at competitive prices, from preferred suppliers, with time to import if necessary, and with no emergency premium.

The Nigerian spare parts supply environment, with its limited local stockholding for many industrial components and its extended import lead times for items not stocked domestically, makes this planned sourcing advantage particularly valuable. A predictive maintenance programme that identifies a developing gearbox bearing fault and estimates that replacement will be needed within six to eight weeks has given the procurement team six to eight weeks to source the correct bearing at the best available price. A reactive maintenance event requiring the same bearing provides the procurement team with the choice between whatever is stocked locally, at whatever price the emergency supplier chooses to charge, or waiting days for delivery from the nearest city. In Nigeria's spare parts environment, the difference between these two sourcing situations, both in price and in downtime duration, is significant and in some cases substantial.

The Skills and Technology Availability Question

A concern that Nigerian manufacturers frequently raise about predictive maintenance is the availability of the skills and technologies needed to implement it effectively in a local context. Vibration analysis expertise, thermal imaging capability, and oil analysis laboratory services are less uniformly available across Nigeria than they are in economies with larger and more developed industrial service sectors. In major industrial centres, Lagos, Port Harcourt, Abuja, Kano, and Ibadan, these services and the expertise to provide them are accessible. In smaller industrial towns, the same resources may require more effort to identify and access.

This concern is legitimate but should not be overstated, for three reasons. First, the most accessible and affordable condition monitoring technologies, including handheld vibration meters, infrared thermometers, and portable thermal imaging cameras, have become considerably less expensive over the past decade and are now purchasable at prices that are within reach of most established Nigerian manufacturing businesses. The barrier to basic condition monitoring technology is lower than it was even five years ago. Second, the oil analysis laboratories that provide the analytical component of lubricant monitoring programmes serve Nigerian industrial customers through a combination of local collection points and courier-based sample submission, making the service geographically accessible even to factories that are not located in major cities. Third, the external condition monitoring service provider model, in which a specialist firm conducts periodic surveys and provides interpreted reports, removes the requirement for internal expertise development entirely for manufacturers who prefer to access the capability as a service rather than build it in-house.

Making the Transition: A Practical Path for Nigerian Manufacturers

Why You Do Not Have to Choose Between the Two

The framing of this article as predictive versus reactive maintenance might suggest that the decision facing Nigerian manufacturers is a binary one: either continue with reactive maintenance as currently practised or make a comprehensive transition to predictive maintenance across all equipment. In reality, the optimal maintenance strategy for most Nigerian manufacturing facilities is a hybrid of all three maintenance approaches: reactive maintenance for low-criticality, easily replaceable equipment; preventive maintenance on a scheduled basis for moderately critical equipment; and predictive, condition-based maintenance for the most critical production assets whose unplanned failure would be most costly and most disruptive.

The journey toward this hybrid model does not require a sudden comprehensive transformation. It is built incrementally, starting with the equipment whose current reactive maintenance approach is causing the most pain and the greatest cost, and progressively expanding the condition monitoring programme as the organisation develops the skills, the baseline data, and the management confidence to extend it. A factory that introduces vibration monitoring on its three most critical rotating machines in the first year, adds thermal imaging surveys of its electrical distribution and motor systems in the second year, and introduces oil analysis for its large gearboxes in the third year has made a steady, affordable transition to genuine predictive maintenance for its most important assets without attempting a transformation that outpaces its organisational capacity.

The First Steps Toward Predictive Capability

For a Nigerian manufacturer who currently operates on reactive maintenance and wants to begin the transition toward predictive capability, the most productive first steps are not necessarily the purchase of technology. They are the establishment of the data disciplines and baseline knowledge that make condition monitoring information meaningful. The first step is completing or refining the asset register and criticality assessment described in the preceding article in this series, identifying clearly which machines are the highest priority candidates for condition monitoring. The second step is establishing a current condition baseline for those priority machines: conducting a thorough inspection, recording the current state of all key components, and taking initial vibration, temperature, and oil sample measurements that will serve as the reference point against which future measurements are compared.

The third step is training: ensuring that the maintenance team members who will conduct condition monitoring surveys understand not just how to operate the measurement equipment but what the measurements mean, what patterns indicate developing faults, and how to communicate their findings to management in terms that connect condition monitoring data to maintenance decisions and production planning. This training investment is the one that most determines whether a predictive maintenance programme delivers on its potential, because the technology captures the data but the trained analyst is the one who converts it into timely, accurate maintenance decisions.

The fourth step is integration with procurement and production planning: establishing the communication channels through which a predictive maintenance finding that a specific component will need replacement within six weeks generates both a purchase order for the replacement part and a discussion with production planning about the best time to schedule the brief planned maintenance window in which the replacement will be made. This integration is where the operational and financial value of predictive maintenance is ultimately realised, in the planned, undramatic, low-cost replacement that prevents the unplanned, destructive, expensive failure that would otherwise have occurred.

Measuring the Return on Predictive Maintenance Investment

Like any significant operational investment, a predictive maintenance programme should be evaluated against a clear set of performance metrics that measure whether it is delivering the returns that justify its cost. The most important metrics for a Nigerian manufacturing facility are the reduction in unplanned downtime hours compared to the pre-programme baseline, the reduction in emergency repair expenditure as a proportion of total maintenance spend, the number of developing faults identified and addressed before failure as a proportion of total maintenance interventions, and the trend in total maintenance cost per unit of production output over time.

These metrics, tracked consistently from the programme's inception, build the evidence base that justifies continued investment in condition monitoring capability and supports the case for expanding the programme to additional equipment categories. They also create the accountability structure that keeps the programme alive and genuinely operational rather than allowing it to drift into a nominal activity that generates data but produces no decisions. A predictive maintenance programme that is regularly evaluated against its own performance targets, whose results are reviewed in management meetings alongside production output and quality data, and whose budget is adjusted based on demonstrated return rather than assumed value is a programme that will continue to improve over time and will remain credible to the business leadership whose continued support it depends on.

Conclusion: The Choice Is Not Between Old and New

Return to Taiwo in Ibadan, contemplating two million, eight hundred thousand naira of avoidable breakdown costs and five failures that his own engineer confirmed were detectable weeks in advance. The question Taiwo is now facing is not whether to move away from pure reactive maintenance. That question was answered by the cost calculation. The question is how far and how fast to move toward condition-based maintenance, and on which equipment to start.

The answer his engineer gave him was specific and practical. Begin with the three roller mills, which together represent the highest proportion of his total breakdown cost and the longest average repair time when they fail. Install a simple monthly vibration monitoring route on the primary bearings and gearboxes of all three machines. Establish temperature baselines for the main motor housings and gearbox cases using an infrared thermometer that already sits unused in the maintenance storeroom. Introduce an oil sampling programme for the roller mill gearboxes, sending quarterly samples to a laboratory in Lagos that offers a two-day turnaround on basic analysis. Do these three things consistently for twelve months and then evaluate the results before deciding what to add next.

This is the appropriate scale and pace for a manufacturer beginning the transition from reactive to predictive maintenance: specific, bounded, affordable, and measurable. Not a comprehensive transformation that overwhelms the organisation's capacity to absorb change, but a targeted first step on the highest-priority equipment, executed with genuine discipline and evaluated honestly against concrete results. The predictive maintenance programmes of Nigeria's best-maintained large factories, the Dangote cement plants with their sophisticated online monitoring systems, the major beverage manufacturers with their planned shutdown programmes calibrated to vibration trend data, all started somewhere comparable to where Taiwo is starting. The distance between his current position and theirs is not the length of a single transformation project. It is the cumulative product of years of consistent, incremental steps in the right direction.

Reactive maintenance will always have a place in Nigerian factories, for the right equipment in the right circumstances. But for every factory's most critical machines, the ones whose failure costs the most in the shortest possible time, the question is not whether predictive maintenance would be better. It is when to start, and the honest answer is that the cost of not starting accumulates with every breakdown that could have been prevented.