Predictive Maintenance: Buying Every Repair Before It Chooses Its Own Moment
Every mechanical failure a fleet has ever suffered was, at some earlier point, a small and affordable problem. A filter that needed changing. A belt showing its first hairline crack. A fault code the driver noticed and decided not to mention. What converts the cheap version of a repair into the tow truck on Mombasa Road, the spoiled cargo, and the engine rebuild is nothing more than information arriving too late to act on. Kendaall’s predictive maintenance service moves that information forward in time: service scheduling driven by true kilometres and true engine hours instead of a driver’s memory, live fault codes and battery health pulled straight from the vehicle’s own electronics, fuel-burn drift and driving-style data read as early wear signals long before a symptom appears, and a per-asset cost history detailed enough to finally answer the question every fleet eventually has to answer honestly — keep this vehicle, or sell it.
Get a Fleet Health Quote Talk Through Your Fleet
Three Ways to Maintain a Fleet. Two of Them Are Just Slower Ways of Paying More.
Every fleet operating in Kenya today runs one of three maintenance regimes, whether or not anyone has ever named it out loud. Reactive maintenance fixes what breaks, after it breaks — the default regime for most operators, and reliably the most expensive one, because a failure that happens in active service is never billed only for the repair itself. It is billed for the recovery vehicle, the lost operating day, the missed delivery slot or the abandoned route, and for whatever damage the original fault caused on its way to becoming a total failure — and for passenger fleets, occasionally for something worse than money. Scheduled maintenance services by a calendar or by an odometer reading someone remembers to check, which is an improvement, but a blind one in both directions at once: the matatu covering six thousand kilometres a month blows straight through its service interval weeks before “quarterly service” ever arrives, while a rarely-used standby pickup receives oil changes it never came close to earning. Predictive maintenance — the regime this service exists to install — schedules and intervenes based on what the asset itself is actually reporting: how far and how hard it has genuinely worked, what its own electronics are flagging in real time, and what its consumption and treatment curves already say is coming before a mechanic would ever be called to look.
This three-way distinction is not a Kendaall invention; it mirrors the language used across industrial maintenance engineering more broadly, where the same three regimes are usually described as reactive (or corrective) maintenance, preventive maintenance, and predictive maintenance, sometimes alongside a fourth category, reliability-centred maintenance, which formalises the process of deciding which strategy suits which failure mode on which asset. Kendaall’s platform is, in effect, a practical, fleet-scale application of that same thinking, built for Kenyan road, agricultural, and plant-hire conditions rather than for a factory floor.
It is worth being precise about what “predictive” actually means in this context, because the word is used loosely elsewhere in the industry. Kendaall’s service is not attempting to forecast the exact date an engine will fail down to the day — no data source available to a road vehicle fleet can promise that level of precision, and any provider claiming otherwise is overstating what usage, diagnostic, consumption, and behavioural data can actually tell you. What the service does responsibly is narrow a wide, unknown risk window into a specific, actionable one: instead of “this vehicle might fail at some unknown point,” the platform says “this vehicle’s brake wear, given its behaviour profile, is likely to need attention within roughly the next two thousand kilometres,” which is a fundamentally more useful and more honest kind of prediction for a fleet to plan around.
The raw material for this shift already flows through the Kendaall platform, because it is the same telemetry the rest of the fleet management account already collects. Every tracked asset reports true distance travelled and — on machinery and generators — true engine hours, so service intervals count themselves automatically instead of relying on a driver’s photograph of a dashboard reading. OBD-connected vehicles stream fault codes, battery voltage, and temperature telemetry, surfacing on the fleet dashboard the warnings that otherwise die quietly on an instrument cluster nobody is watching. Fuel monitoring contributes the subtlest signal of all — consumption drift, the thirsty engine announcing an injector or air-filter problem months before a workshop would find it unprompted — and driver behaviour data explains the wear-rate gap between two identical vehicles, so brake and clutch forecasting can follow how a specific vehicle has actually been treated rather than a fleet-wide average that flatters some vehicles and understates risk on others. Predictive maintenance is the layer where all four streams converge into a single, specific instruction: this vehicle, this system, this week — before it chooses its own moment.
The repair that happens in the yard on a Tuesday instead of on the highway at midnight. (Replace with your workshop photo.)
The economics compound across an asset’s entire working life rather than showing up in any single invoice. Downtime is the multiplier most operators undercount, because it rarely appears as a line item anywhere: a truck earning revenue daily that spends three days off the road for an emergency repair has lost more in foregone income than most repairs actually cost to perform; a matatu pulled off its route surrenders that route, and its regular passengers, to a competitor for as long as it is out of service; a harvester dead in week two of a harvest costs a weather window that no invoice line can ever fully express. The quieter payoff arrives at the far end of each asset’s life, when the vehicle carrying a complete, dated, telemetry-backed service history sells for more, defends its warranty position more credibly, and lets the keep-or-sell decision be made against a cost-per-kilometre curve rather than a gut feeling in the yard. The service runs fleet-wide across everything already on the platform: logistics trucks, corporate vehicles, hire cars ageing rapidly in renters’ hands, ride-hailing units compressing a private car’s typical lifetime mileage into a single year, plant, tractors, and gensets — one maintenance brain sitting over the whole fleet management estate rather than a separate system for each asset class.
Where Your Fleet Is Now, and What Each Step Up Is Worth
Most operators, when they look honestly at their own operation, can place themselves on this scale without much difficulty. The value of naming the regime explicitly is that it turns an abstract goal — “better maintenance” — into a specific, budgetable move from one named state to the next.
Reactive: Fix What Breaks
Repairs at highway prices plus downtime, recovery, and collateral damage. The regime most fleets are in, and the one none chose on purpose. Its true cost hides in lost operating days, not workshop invoices, which is why it is so often underestimated by owners who only ever look at the repair bill itself.
Scheduled: Fix by Calendar
Better — but calendars don’t know that one truck did double the kilometres and the other barely moved. High-use assets slip past their real intervals while idle assets get serviced for nothing. Precision is spent where it isn’t needed and missing exactly where it matters most.
Predictive: Fix What the Data Names
Intervals counted by true use, faults surfaced as they register, wear signals read from fuel and driving curves — the repair bought early, in the yard, on a day the fleet chose rather than a day the vehicle imposed. This is the regime this service installs.
Four Kinds of Signal, and Which Failures Each One Catches
Predictive maintenance is sometimes described as if it were a single technique, but in practice it is closer to a family of related signal types, each suited to a different category of failure. Understanding the differences helps explain why the service is built as five separate modules rather than one generic “health score.”
Usage Signals
The simplest and most reliable category: how far or how long an asset has actually operated since its last service. Usage signals catch the entire class of wear-based failures — oil breakdown, filter saturation, belt fatigue, brake pad wear — that occur predictably as a function of distance or hours rather than calendar time. They require no special sensor beyond accurate GPS distance or an ignition-wired hour meter, which is why they form the foundation of the service.
Diagnostic Signals
Electronic self-reporting from the vehicle itself, in the form of OBD-II and CAN bus fault codes, battery voltage, and temperature readings. Diagnostic signals catch failures the vehicle’s own engine control unit is already capable of detecting internally — a misfire, a sensor out of range, a battery losing capacity — but which typically only reach a human via a dashboard warning light that a driver can, and often does, ignore or fail to report.
Consumption Signals
Trends in fuel burn rate, benchmarked per asset against its own historical baseline or against similar vehicles doing similar work. Consumption signals catch a category of mechanical degradation — clogged injectors, failing turbochargers, dragging brakes, under-inflated tyres — that often produces no fault code and no obvious symptom at all until it has become expensive, because the engine compensates for the underlying problem for a surprisingly long time before performance visibly suffers.
Behavioural Signals
Data describing how a vehicle is driven — harsh braking, harsh acceleration, cornering forces, idling time — used to forecast wear on components that degrade proportionally to treatment rather than to distance alone. Two vehicles covering identical mileage can have very different real remaining life in their brakes and clutches, and behavioural signals are what allow the platform to tell them apart before the wear becomes a failure.
None of these four signal types is sufficient on its own. Usage signals alone would still service every vehicle identically regardless of how gently or harshly it was driven; diagnostic signals alone would miss the slow degradation that never triggers a formal fault code; consumption signals alone would say something was wrong without saying what; and behavioural signals alone would say nothing about the vehicle’s actual mechanical condition. The service is built to read all four together, on the understanding that a genuinely predictive picture of an asset’s health is a combination of what it has done, what it is reporting, what it is consuming, and how it has been treated — not any single one of those in isolation.
It is also worth noting that these four signal types accumulate value at different rates over time. Usage signals are useful from the very first day a service interval is defined, since a kilometre counted today is exactly as reliable as a kilometre counted a year from now. Diagnostic signals are similarly immediate, since a fault code means the same thing the first time it appears as it ever will. Consumption and behavioural signals, by contrast, only become genuinely predictive once a meaningful baseline has been established for each specific asset, which is why the platform treats the first few weeks after installation as a calibration period rather than expecting drift and wear alerts to be fully accurate from the moment a device is fitted.
Six Ways Fleets Pay Twice for Maintenance — and the Signal That Prevents Each
The Breakdown at the Worst Possible Kilometre
Failures don’t schedule themselves around your operations — they compound quietly and then choose the loaded trip, the school run, the harvest week. Usage-counted intervals, live fault codes, and health flags move the intervention weeks earlier, converting the roadside emergency into a yard appointment. The repair is often literally the same part; only the price, the downtime, and the danger change.
Intervals Managed by Memory and Driver Photos
“When was KCF 402D last serviced?” should not be a research project. True-distance and true-hour counting per asset makes every interval self-tracking: approaching-service alerts to the workshop desk, overdue flags that escalate, and — for the assets that must not gamble — “do not dispatch” holds visible wherever vehicles are assigned. The ride-hailing unit doing 5,000 km a month hits its interval every four weeks; now something other than the driver’s honesty says so.
The Fault Code That Died on the Dashboard
Vehicles announce their problems constantly — to drivers, who have a route to finish and a light they’ve learned to ignore. OBD telemetry routes the announcement past the dashboard to the fleet desk: fault codes with descriptions, battery voltage sagging toward the no-start morning, temperature anomalies flagged while they’re still a coolant issue and not a head gasket. The driver’s silence stops being the fleet’s blindness.
Identical Vehicles, Mysterious Cost Gaps
Two matching trucks, one costing half again as much to keep alive — the gap has causes, and the platform names them: the driving profile writing the brake bill, the consumption drift flagging the injectors, the route whose corrugations show up in suspension spend. Wear forecasting follows treatment instead of fleet averages, so parts are staged for the vehicle that will actually need them first.
Servicing the Fleet Instead of the Asset
Blanket maintenance wastes precision in both directions — the idle asset over-serviced by calendar, the hard-run one starved by the same calendar. Per-asset scheduling spends the workshop budget where the use went: the harvest-season tractor serviced by its thousand hours, the standby genset by its seventy, and neither by the quarter. Fleets typically find the reallocation alone funds the service.
Keep-or-Sell Decided by Sentiment
Every fleet has the vehicle everyone defends and the spreadsheet no one keeps. Per-asset lifecycle records — cost per kilometre trending over time, downtime days, repair frequency — draw the curve that shows exactly when maintenance spend crossed replacement logic, and the complete telemetry-backed service history adds real money at resale. Renewal decisions move from the loudest opinion in the yard to a chart both the accountant and the mechanic accept.
Five Modules From Signal to Decision
Each module below corresponds to one stage in the path from raw telemetry to a decision a fleet manager or owner can actually act on. Most fleets adopt all five together, since each module strengthens the accuracy of the others, but they can also be enabled incrementally, starting with usage-based scheduling and adding the diagnostic, consumption, and lifecycle layers as budget and hardware allow.
MODULE 01 · Usage-Based Service Scheduling
Every asset’s intervals — oil, filters, brakes, belts, tyres, inspections — are defined once and counted automatically against true GPS distance or ignition-wired engine hours. Approaching-threshold alerts reach the workshop desk with enough lead time to plan around them; overdue items escalate through the chain of responsibility; and safety-critical intervals can impose dispatch holds visible at the exact point vehicles are assigned to a job. Calendars remain available for the small number of items that genuinely age by time rather than by use — a rubber seal, a coolant additive with a shelf life — while everything else follows actual use.
- Per-asset interval plans on true kilometres and hours
- Approaching, due, and overdue alerts with escalation
- Dispatch holds for safety-critical overdue items
- Season-readiness reports for agricultural and project fleets
MODULE 02 · Vehicle Health Telemetry & Fault Codes
Where the vehicle offers a diagnostic interface, the platform listens continuously: OBD-II and CAN fault codes decoded and routed to the fleet desk in plain language rather than raw hexadecimal codes, battery voltage trended against the no-start threshold, coolant and operating temperatures watched for the gradual drift that precedes a boil-over. Where older vehicles and machines offer nothing electronic to read, our own sensing supplies the essentials — the same instrument-over-electronics philosophy our plant clients already rely on across mixed-age fleets that span several decades of vehicle manufacture.
- Live fault-code capture with plain-language descriptions
- Battery health trending and pre-failure alerts
- Temperature anomaly detection on engines and reefer units
- Coverage across modern OBD vehicles and bare mechanical assets
MODULE 03 · Early Wear Signals: Fuel Drift & Treatment Data
The predictive layer proper, and arguably the module with the least intuitive payoff and the largest actual return. Consumption benchmarked per asset class turns fuel-burn drift into a mechanical early warning — an engine drinking eight percent above its twin is naming a problem months ahead of any driver-noticeable symptom — while harsh-event histories forecast brake, clutch, tyre, and suspension wear according to how each vehicle is actually treated rather than fleet-wide averages that flatter careful drivers’ vehicles and understate risk on the rest. The two curves together are why identical vehicles stop having mysterious, unexplained cost gaps between them.
- Consumption-drift alerts as engine-health signals
- Treatment-weighted wear forecasting per vehicle
- Route-severity factors for corrugation and terrain wear
- Parts staging suggestions ordered by predicted need
MODULE 04 · Workshop Workflow & Service Records
The operational spine of the whole service: every service event logged against the correct asset with the date, the true reading at the time, the work performed, the parts used, and the cost — building a complete history automatically as the fleet runs, rather than as a separate administrative task someone has to remember to do. Reminders route to the right desk, completed work closes its interval and restarts the count immediately, and the record travels with the asset through ownership changes. For fleets that use external garages, the exportable job card and history keep every workshop working from the same set of facts, regardless of which one performed the previous service.
- Per-asset service logs with readings, work, parts, and cost
- Interval auto-reset on completed work
- Exportable job cards and histories for external workshops
- Warranty-grade documentation, dated and telemetry-backed
MODULE 05 · Lifecycle Economics: Cost Curves & Keep-or-Sell
The decision layer, and the one that turns maintenance data into a strategic input rather than an operational one. Cost per kilometre (or per hour, for machinery and gensets) is trended over each asset’s entire working life, downtime days are counted precisely, and repair-frequency curves are drawn — letting the renewal question be answered at the exact point where the lines cross rather than by instinct. Fleet-level views rank the whole estate from cheapest-to-run to quietly ruinous, informing which units to duplicate on the next purchase, which to retire now, and what a disposal listing can honestly claim, with the full service history acting as the proof that measurably lifts the sale price.
- Cost-per-km / cost-per-hour lifecycle curves per asset
- Downtime and repair-frequency tracking
- Fleet ranking for renewal and disposal planning
- Resale-ready service history exports
The fleet’s health on one screen: intervals counting themselves, faults surfacing, and the cost curve that ends the keep-or-sell debate. (Replace with a dashboard screenshot.)
The Vocabulary Behind Predictive Maintenance, in One Place
Maintenance conversations often stall because a handful of terms get used loosely, interchangeably, or slightly differently by mechanics, fleet managers, and software vendors. The definitions below are how Kendaall uses each term across the platform and this page.
Service Interval
The usage threshold — expressed in kilometres, engine hours, or occasionally calendar time — at which a specific maintenance task (an oil change, a filter replacement, a brake inspection) becomes due. Different components on the same vehicle typically have entirely different intervals running in parallel.
OBD-II / CAN Bus
On-board diagnostics is the standardised system built into most vehicles since the late 1990s and 2000s that reports fault codes and live sensor data through a common connector, carried internally over a controller area network (CAN) bus. It is the interface Kendaall’s telemetry hardware reads from to capture fault codes, battery voltage, and temperature data without any modification to the vehicle itself.
Fault Code
A standardised alphanumeric code generated by a vehicle’s engine control unit when a monitored system falls outside its expected operating range — a lean fuel mixture, a sensor reading out of bounds, a misfire on a specific cylinder. Fault codes are the vehicle’s own diagnosis, decoded and routed to a human rather than left to blink on a dashboard.
Consumption Drift
A gradual, sustained change in a vehicle’s fuel burn per kilometre or per hour relative to its own historical baseline, measured independently of short-term variation caused by traffic, load, or terrain. Drift is treated as a mechanical signal in its own right, distinct from the theft-detection role fuel monitoring plays elsewhere on the platform.
Wear Forecasting
The practice of predicting a component’s remaining useful life based on how a specific asset has actually been operated — braking intensity, cornering forces, load carried — rather than applying a single generic mileage-based interval to every vehicle in a class regardless of how differently each one is driven.
Lifecycle Cost Curve
A trend line, usually cost per kilometre or cost per hour, tracking an asset’s total maintenance and repair spend against its use over its entire working life. The point at which this curve begins rising sharply and consistently is generally the point at which replacement becomes more economical than continued repair.
Dispatch Hold
A flag applied to an asset whose safety-critical service interval has been exceeded, preventing it from being assigned to a job until the required maintenance is recorded as complete. Dispatch holds exist specifically to stop an overdue interval from being quietly worked around under operational pressure.
Predictive Maintenance (as distinct from Preventive)
Preventive maintenance acts on a fixed schedule regardless of an asset’s actual condition; predictive maintenance acts on evidence of the asset’s actual condition, whether that evidence is usage-based, diagnostic, consumption-based, or behavioural. In practice the two overlap heavily — a usage-counted interval is arguably still a form of preventive maintenance, just a far more accurate one — and Kendaall’s service is best understood as a spectrum running from accurate preventive scheduling through to genuinely condition-based predictive intervention, rather than a single hard line between the two.
The Same Five Modules, Weighted Differently Per Asset Class
Predictive maintenance is not a single fixed configuration applied uniformly to every vehicle on an account. The five modules described above are the same everywhere, but which ones carry the most weight, and which failure modes matter most, changes meaningfully depending on what kind of asset is being monitored and how it earns its keep.
Trucks and Logistics Vehicles
For long-haul and delivery trucks, usage-based scheduling and consumption drift dominate, because these vehicles accumulate distance quickly and unevenly across a fleet, and fuel is typically the single largest controllable operating cost after the vehicle itself. Behavioural wear forecasting matters heavily too, since braking and cornering discipline on loaded trucks has an outsized effect on brake and tyre life. See how this combines with route and delivery data on the logistics fleet tracking page.
PSVs and Ride-Hailing Vehicles
Passenger service vehicles and ride-hailing cars compress a private car’s typical lifetime mileage into a single year or two, making usage-based intervals essential simply to keep pace with real wear. Diagnostic signals carry particular weight here as well, because a fault ignored by a driver under pressure to keep earning has direct safety implications for passengers, detailed further on the PSV fleet tracking and taxi and ride-hailing pages.
Heavy Machinery and Plant
Excavators, graders, and loaders are billed and serviced by engine hours rather than distance, and their consumption patterns vary enormously with the type of work being performed on any given day. Usage-based scheduling by hours and consumption-drift monitoring are the dominant modules, paired with the theft-protection and utilisation data covered on the heavy machinery tracking page.
Generators
Standby and prime-power generators live or die, reputationally, on a single mains-failure event, so diagnostic and usage signals matter less than the mains-sense and changeover data covered in depth on the generator tracking page — with predictive maintenance layered on top to schedule servicing by true runtime hours rather than an under-used or over-used calendar interval.
Agricultural Equipment
Tractors and harvesters see extremely concentrated use during planting and harvest windows and near-total idleness the rest of the year, making usage-based, season-readiness scheduling particularly valuable — an equipment failure during a narrow harvest window carries a cost that has little to do with the repair bill itself, as explored on the agriculture tracking page.
Hired-Out Vehicles and Equipment
Assets on hire are, by definition, operated by someone other than their owner, which makes behavioural and consumption signals especially valuable as an early warning that a renter is treating an asset harshly, alongside the billing and inventory logic covered on the car hire fleet tracking page.
How a Fleet Actually Moves From Reactive to Predictive
The shift described throughout this page rarely happens in one step, and it doesn’t need to. Fleets that adopt predictive maintenance successfully tend to follow a broadly similar sequence, regardless of size or asset mix, and understanding that sequence in advance makes the transition considerably less disruptive than owners often expect.
Step One: Baseline the Fleet
Before any predictive signal means anything, the platform needs a true picture of what “normal” looks like for each asset — typical fuel consumption, typical driving patterns, current service status. This baseline period, usually two to four weeks, is when the usage-based scheduling module is configured against manufacturer or fleet-specific intervals and existing service records are loaded in wherever they exist.
Step Two: Turn On Diagnostic and Consumption Layers
Once the baseline is established, fault-code monitoring, battery health tracking, and consumption-drift alerts are switched on. This is typically the point at which a fleet receives its first genuinely surprising alert — a vehicle whose fuel burn has been quietly drifting for months, or a battery already trending toward a Monday-morning no-start — and where the value of the service becomes concrete rather than theoretical.
Step Three: Layer In Behavioural Wear Forecasting
With driving-behaviour data accumulating in parallel, wear forecasting begins to separate vehicles that looked identical on paper into genuinely different risk profiles, and parts staging can start following predicted need rather than a single fleet-wide reorder point.
Step Four: Build the Lifecycle View
As service records, downtime events, and repair costs accumulate against each asset over subsequent months, the lifecycle cost curves described in Module Five begin to take a meaningful shape, and the keep-or-sell conversation moves from opinion to evidence for the first time.
Most fleets are running the full five-module regime within one full service cycle of onboarding, and the modules can be enabled in a different order than described here if a fleet has a more urgent priority — a hire company anxious about billing disputes, for instance, might prioritise the workshop and lifecycle modules earlier, while a hospital or telecom operator might prioritise diagnostic and mains-related signals from day one.
Where the Numbers Actually Move: Downtime, Repair Cost, and Resale Value
Fleet owners evaluating predictive maintenance tend to ask a version of the same question: where, specifically, does the subscription pay for itself? Unlike some fleet technologies whose value is diffuse or hard to isolate, predictive maintenance concentrates its return in three measurable places, and most operators can validate the business case against their own historical numbers before committing to anything.
The first and largest is downtime avoided. A breakdown in active service costs far more than the parts and labour that eventually appear on the invoice, because the vehicle earns nothing while it sits in a workshop or on the roadside, and because a recovery operation, a missed delivery window, or a cancelled route day each carry their own separate cost that rarely gets attributed back to the maintenance decision that could have prevented it. Fleets that track downtime days before and after adopting usage-based scheduling and fault-code monitoring typically see the clearest, fastest-arriving evidence of the service paying for itself, often within the first one or two service cycles.
The second is the cost differential between a planned repair and an emergency one. The same worn component, replaced on a scheduled workshop visit versus replaced after it fails on the road, is frequently the same part at a very different total cost — emergency labour rates, recovery fees, and consequential damage from a component that failed catastrophically rather than being caught while merely degraded. Consumption-drift and behavioural wear signals exist specifically to move a repair from the second category into the first.
The third, and the one most consistently underestimated by owners focused on day-to-day running costs, is resale value. A vehicle sold with a complete, dated, telemetry-backed service history is a fundamentally different proposition to a buyer than one sold on the seller’s word alone, and serious used-vehicle and used-equipment buyers routinely discount their offers to account for exactly the uncertainty that a verified history removes. Fleets that export Kendaall service histories at the point of sale consistently report both faster sales and firmer negotiated prices, turning a maintenance record kept for operational reasons into a direct contributor to the fleet’s eventual disposal proceeds.
A fourth, less frequently discussed return sits in staff time rather than cash. Fleet managers running a reactive or loosely scheduled operation spend a disproportionate amount of their week firefighting — chasing down which vehicle is where, whether it was serviced, and why it has broken down again — time that is difficult to value precisely but is nonetheless real and recoverable. Once intervals, faults, and history are tracked automatically rather than chased manually, that time typically shifts toward planning and toward the kind of proactive supplier and workshop negotiation that a constantly reactive operation rarely has the bandwidth for.
What Changes in the First Service Cycle
| Question | Before Predictive Maintenance | After Predictive Maintenance |
|---|---|---|
| When is this vehicle’s next service due? | Estimated from memory or the driver’s last odometer photo | Counted automatically against true GPS kilometres or engine hours |
| Is there a developing fault right now? | Unknown until a dashboard light is noticed and reported, if ever | Fault codes and battery/temperature anomalies routed to the fleet desk instantly |
| Why does this vehicle cost more to run than its twin? | Put down to “bad luck” or an unverified guess | Traced to specific driving behaviour or consumption drift |
| Is this vehicle worth keeping? | Decided by opinion in the yard | Decided against a cost-per-kilometre curve trending over its life |
| What is this vehicle’s service history at resale? | Whatever paperwork survived, if any | A complete, dated, exportable, telemetry-backed record |
What Fleet Owners Ask Before Moving to Predictive
Does this work on our older vehicles, or only new ones with OBD ports?
Both, at different depths. Modern vehicles contribute the full telemetry layer — fault codes, battery voltage, temperatures — through their diagnostic interfaces. Older vehicles and bare mechanical machines still get the regime’s core: true-distance and true-hour interval counting, fuel-drift signals where a sensor is fitted, behaviour-based wear data, and the complete service record. In practice the usage-counted intervals alone capture most of the value, and they work on a 1998 canter as well as a 2024 truck.
We already have a service schedule. What does this add?
Three things your calendar can’t do: it counts actual use per asset (so the hard-run truck isn’t serviced on the idle one’s timetable), it hears what vehicles report between services (the fault code, the sagging battery, the thirsty week), and it remembers everything in a history that survives staff changes and garage changes. Most fleets keep their existing garage relationships entirely — the service changes when and why vehicles arrive at the workshop, not where.
Can it really predict a failure before it happens?
Honestly framed: it predicts risk windows, not appointments with destiny. A drifting consumption curve, a rising fault-code pattern, a battery trending toward its threshold, an interval blown past — each names a system and a timeframe in which intervention is cheap and inaction historically hasn’t been. Fleets that act on the flags see breakdown rates fall steeply; the failures that remain tend to be the genuinely unpredictable minority. What disappears almost entirely is the category of breakdown that was announced and ignored.
Who receives the alerts — we don’t have a maintenance department.
The routing fits the fleet: a three-car operation sends everything to the owner’s phone in plain language (“KCD 118B: service due within 400 km”); larger fleets route by role — workshop desk for intervals, operations for holds, management for the monthly summary. Fleets without any workshop arrangement can pair the flags with our partner-garage referrals, and device-side issues route automatically to the repair & maintenance team so the monitoring layer maintains itself too.
What does it cost, and where’s the payback?
For fleets already on Kendaall tracking, predictive maintenance is a platform capability — enabled per asset at a modest addition detailed on the pricing page — with OBD leads or fuel sensors as the only hardware where those layers are wanted. Payback arrives as absence: the breakdown that didn’t happen, the engine that got injectors instead of a rebuild, the downtime days that stayed operating days. The measurable version most clients cite is downtime — the metric worth tracking for one quarter before and after; request a fleet-specific model through the quote form.
Does the service history really affect resale value?
Ask any serious used-vehicle buyer what they trust least: the seller’s memory. A dated, complete, telemetry-backed history — every service at its true reading, every fault addressed, downtime visible — is the difference between a price defended with documents and a price negotiated against suspicion. Fleets disposing of vehicles with exported Kendaall histories consistently report faster sales and firmer prices, and the same file carries the vehicle through ownership transfer cleanly.
Can predictive maintenance actually reduce our insurance costs or claims?
Indirectly, and increasingly so. A vehicle with a documented service history and evidence of proactive fault resolution is a stronger position in any claim involving a mechanical contributing factor, and some insurers already offer preferential terms to fleets that can demonstrate structured maintenance practices. While Kendaall does not set insurance policy, the exportable service and fault history is exactly the documentation insurers and claims assessors ask for after an incident, and fleets have used it successfully to support their position.
How does this fit with reliability-centred maintenance principles used in larger industrial fleets?
Very directly. Reliability-centred maintenance asks, for each component, which failure modes matter most and which strategy — reactive, scheduled, or condition-based — suits each one, rather than applying one blanket approach fleet-wide. Kendaall’s four signal types (usage, diagnostic, consumption, behavioural) map onto exactly that thinking: safety-critical items get usage-based holds, electronically-monitored systems get diagnostic alerts, and slow-degrading components get consumption and behavioural forecasting, matched to the failure mode rather than forced into a single generic schedule.
Will drivers or workshop staff resist this kind of monitoring?
Some initial resistance is normal, and usually short-lived once the practical benefit becomes obvious. Drivers who previously carried informal responsibility for noticing and reporting problems tend to appreciate a system that catches issues they might otherwise be blamed for missing, and workshop staff generally welcome clear, prioritised job lists over a queue of vehicles arriving unannounced with vague complaints. Framing the rollout around reducing roadside emergencies — which are stressful and inconvenient for drivers and mechanics alike — tends to build buy-in faster than framing it as oversight.
Buy Your Next Ten Repairs at Yard Prices Instead of Highway Prices
Send us the fleet list and how it works — routes, duty cycles, current service arrangement — and we’ll return the maintenance configuration per asset class, the hardware (if any) it needs, and a downtime-based payback model against your own operation.
Request a Fleet Health Quote Hear From Fleet Clients