In this article, you'll discover where repair time is really lost, what those delays cost OEMs and how better product knowledge can shorten the journey back to operation.
When a complex machine fails, attention naturally turns to the repair itself. A technician diagnoses the fault, removes the failed component, installs the replacement and tests the machine.
That is the visible part of the job, but it's rarely the whole story.
Before the technician can complete the repair, somebody may need to confirm which machine configuration they are dealing with. They may search through manuals, wait for technical support, compare several similar components, check whether a part number has been superseded and find out where the replacement is stocked.
Once the correct part arrives, they may still need to find the right procedure, clarify an ambiguous instruction, locate a specialist tool or ask a more experienced colleague for help.
The physical repair may take two hours - but the machine could remain unavailable for three days.
All of those activities surrounding the repair contributes to Non-Active Repair Time (NART). For OEMs and their customers, this is often where the greatest opportunity to reduce downtime is hiding.
Non-active repair time is the portion of a repair journey during which the product remains unavailable, but no visible, productive progress is being made towards restoring it.
A simple way to represent the journey is:
Total time to restore operation = active repair time + non-active repair time
Active repair time normally includes activities such as:
In contrast, non-active repair time includes:
Don't get me wrong, not all downtime is bad, and some of the non-active repair time is necessary. Technicians must work safely, parts need to travel and certain repairs require specialist skills. The objective therefore, is to eliminate avoidable NART and compress the delays that cannot be removed completely.
There is no universal active-to-non-active repair ratio that applies to every machine, industry or service model, although many sources cite anywhere between 50% and 80% of the repair time being 'non-active'.
A technician repairing a standard product with the correct part on the van may spend most of the incident making productive progress, but a technician supporting a highly configured machine in a remote location may spend days waiting for information, access, travel or parts before completing a relatively simple physical repair.
In oil & gas, medical and military equipment with a tethered workforce and remote locations, some OEMs have quoted as little as 5% of repair time being active which gives a NART-to-active ratio of 19:1. Such a ratio is entirely possible when a repair involves long logistics delays or several failed attempts, but more commonly the 20-50% statistic is what OEMs in other industries commonly estimate.
Although not an exact match, research into wrench time gives us a useful indication of how much paid technician time can be consumed by surrounding activities. Wrench time looks at the proportion of an available shift during which technicians are making productive progress on assigned work, with maintenance planning specialist Doc Palmer, citing around 35% wrench time for a typical maintenance workforce and approximately 55% as best practice. The remaining time can include travel, waiting for parts, tools, instructions, clearances and coordination. Palmer also warns that the metric can become misleading if organisations use different definitions or reward people for appearing busy rather than completing the correct work. His explanation of wrench time is an important reminder that productive activity and successful repair outcomes are not always the same thing.
Wrench time should not be substituted for NART however. One measures technician activity across a working period, while the other measures delay within an end-to-end repair journey. However, both reveal how many apparently small interruptions can accumulate around the work.
The most useful ratio is therefore the one calculated from your own service events:
NART % = Non-Active Repair Time ÷ Total Time to Restore Operation × 100
NART-to-Active Ratio = Non-Active Repair Hours ÷ Active Repair Hours
For example, if a machine is unavailable for 20 hours and four of those hours are spent actively diagnosing, repairing and testing it:
Using these numbers, we can see that reducing the active repair from four hours to three would save one hour. However, removing a six-hour delay in finding and delivering the correct part would have a much larger effect!
NART rarely comes from one dramatic failure in the service process. Instead, it's usually the compound effect of a series of handovers, searches, uncertainties and queues that relate to the fact that product knowledge doesn't get into the hands of the people that need it fast enough.
| Stage of the journey | Where time is lost |
|---|---|
| Fault reporting | Incomplete symptoms, missing photographs, no serial number, unclear machine location or delayed escalation. |
| Understanding the product | Teams cannot quickly establish the as-built or as-maintained configuration of the machine. |
| Diagnosis | Fault codes lack context, service history is unavailable or previous resolutions cannot be found. |
| Accessing knowledge | Information is spread across PDFs, CAD files, spreadsheets, old catalogues and individual experts. |
| Identifying the part | Similar components, configuration differences, unclear illustrations and missing fitment rules create uncertainty. |
| Confirming the solution | Technicians wait for engineering, warranty, parts or technical support teams to validate the decision. |
| Ordering and logistics | Availability is unclear, the order is entered manually, the wrong part is supplied or emergency freight is required. |
| Performing the work | Instructions are outdated, too generic or difficult to follow in the flow of work. |
| Testing and return | Acceptance criteria are unclear, testing must be repeated or another fault appears. |
| Closing the incident | The resolution is recorded as free text or not captured, so the next technician repeats the same investigation. |
Each individual delay can look insignificant when viewed inside one department, but the end customer relying on the product perceives the entire combined effect as downtime.
Here are some examples of how NART can compound when product knowledge is trapped or difficult to access:
“Model X” is rarely enough information to support a complex product.
Engine type, software version, production date, optional packages, regional specifications and previous modifications can all change the correct diagnosis, part or procedure. If support teams cannot see the product as it exists in the field, every subsequent decision carries more risk.
DRVEN's Founder and CEO, Sam Burgess knows the issue all too well as an operator in the British Army:
"I can still remember sitting in a field, getting wet and rained on with a stack of technical manuals, trying to find the matching parts and the right instructions I needed to repair the equipment. But the physical PDF document bore little resemblance to the actual asset, making the process of repairing it feel like deciphering an ancient code. I knew there had to be a better way."
99% of the time, OEMs have all of the information required to support the machine - the problem is finding and interpreting it quickly.
Engineering may hold CAD and BoM data, parts teams work in an ERP, technical authors manage manuals, service history sits in another system, while practical knowledge is retained by a handful of experienced engineers.
A component can be visually similar to several alternatives. It may have been replaced by a newer part number, apply only to a particular configuration or require an associated kit.
If a parts catalogue cannot guide the user through those relationships, the technician sends photographs, compares dimensions or asks support to confirm the choice. A wrong decision creates even more NART through returns, repeat ordering and another visit to the machine.
Most OEMs have someone who knows exactly what to check when a particular product fails. That experience is extremely valuable, but relying on one person to interpret every difficult case creates a queue.
Support teams become overloaded, less experienced technicians wait for answers, the expert is repeatedly interrupted, and has less time to solve the complex problems that genuinely need their involvement.
A technically correct instruction can still be difficult to use.
It may cover the wrong variant, reference an old part, assume knowledge the technician does not have, or describe a three-dimensional task using an overly dense block of text that's hard to follow. Technicians then pause, improvise or call for clarification - more NART!
If the procedure then leads to an incorrect repair, the resulting repeat visit creates another complete cycle of diagnosis, parts identification and logistics - and yep, more NART.
Some delays sit outside the product knowledge. For example, the required component or spare part may be out of stock, a permit may be required, or a specialist tool may be unavailable or being used at another site.
However, better information can still shorten these stages because if the probable repair, required parts, tools and safety steps are understood much earlier, they can be prepared before the technician even reaches the machine.
The largest financial impact may initially appear on the customer’s side because their production has stopped, an operator cannot complete a job, a vehicle is unavailable, or a contractual commitment is at high risk.
A 2024 Siemens study of large manufacturing and industrial organisations estimated that an hour of unplanned downtime ranged from $36,000 in fast-moving consumer goods to $2.3 million in automotive manufacturing. These figures highlight how quickly the customer’s exposure can grow.
The same study found that the average time required to recover from a downtime incident had increased from 49 to 81 minutes over five years, with skills shortages, supply-chain delays and more complex product failures among the explanations. The True Cost of Downtime 2024 sets out the research and its methodology.
For the OEM, the cost appears in several other places too:
Support engineers spend time answering avoidable questions, locating drawings, interpreting photographs and confirming part numbers. Similarly, field engineers make repeat journeys, parts teams process returns, and emergency freight and overtime become accepted as normal ways of recovering from delays.
These costs are then spread across different budgets, which makes the total cost of NART difficult to see as a single commercial problem - and ultimately lacks a dedicated person or team to focus on all the different ways NART can be reduced.
An unclear repair journey can increase warranty labour, travel, replacement parts and goodwill payments. Service-level agreements or availability-based contracts may also face credits and penalties when restoration targets are missed; leading to OEMs overcompensating by shipping 12 machines instead of 10, or putting technicians on site 24/7 often twiddling their thumbs waiting for the moment of failure that may or may not come.
When the OEM experience is slow or confusing, customers and dealers begin looking for easier alternatives. They may source a third-party unofficial part, use a local repair provider, keep random PDFs of old information - or worse still could start diagnosing via Reddit or YouTube.
The immediate repair may eventually be completed, but part of the ongoing aftermarket relationship has been lost, and if using unofficial parts or instructions they could create further damage or impact the product in unknown ways.
Every avoidable escalation occupies someone who could be handling another customer or supporting a more difficult issue. NART therefore positively affects the capacity of technical support, field service, warranty and parts teams.
As one VP or Aftermarket Sales in the Industrial Machinery sector told DRVEN:
"Our electronic parts catalogue was too confusing and cumbersome for our technicians and dealers to use. This led to our support team overloaded with calls. By implementing DRVEN, technicians and dealers can self-serve saving 1,700 hours of reduced support call hours!"
Customers remember what happened after the machine stopped, so if they experienced uncertainty, repeated visits and long waits for answers, that experience influences renewals, future purchases and their willingness to recommend the OEM.
A useful way to build the business case to focus on NART reduction is:
Annual NART Exposure = Avoidable NART Hours × Relevant Cost Per Hour, plus avoidable support, logistics, warranty and rework costs
The cost per hour should reflect your commercial model.
For the customer, it may include lost production, idle labour, missed deliveries or replacement equipment, and for the OEM it may include engineering time, field labour, freight, returns, warranty cost and contractual penalties.
Keep the calculations separate to avoid double counting, then show how the same delay affects both sides of the relationship and why a product knowledge hub investment to reduce NART therefore makes sense for the business.
Reducing NART starts by making the machine easier to understand, so at the point of need, the technician is able to establish:
This requires more than digitising an existing manual into PDFs. Turning a printed page or PDF into another PDF does not resolve the relationships between product configuration, parts, procedures, engineering changes and service history.
Instead, the product knowledge needs to be structured, connected and delivered in the flow of work.
For a technician, this may mean selecting a component directly from an interactive 3D model and moving immediately into the correct repair procedure. For a dealer, it may mean confirming fitment and availability without contacting the OEM directly. And for a support engineer, it may mean seeing the same machine configuration and service information as the person in the field.
The result is that everyone works from the same understanding of the product, even though they use it for different decisions.
Some service delays cannot be removed completely: A technician may still need to travel, a safe isolation process still needs to be followed, and a replacement part must still need time to reach the site.
Better product knowledge changes when those activities begin and how often they need to be repeated.
If connected machine data indicates a likely failure before arrival, then the technician can bring the relevant parts and tools, and if the exact product configuration is known, the parts team can avoid several rounds of confirmation. Similarly, if a visual procedure includes the isolation, repair and verification steps, the work can begin with greater confidence, and removes the uncertainty, searching and rework preventing skilled people from completing the job safely and correctly.
NART is often treated as an aftermarket process problem, but many of its causes were introduced much earlier where the OEM is not designing their products with supportability in mind.
For example, a component positioned behind several unrelated assemblies will take longer to access, and a fault that cannot be isolated creates more diagnostic work. That may have been ok for production but creates pain for the support & maintenance teams who are on the hook to repair the same awkward components all the time.
This is where non-active repair time connects to designing for supportability.
For an OEM, designing for supportability can include:
The last point is essential, because product knowledge should not travel in only one direction to the aftermarket teams, but also close the loop by providing essential data about faults and fixes that will help design more easily repairable products in the future.
Every service or support event contains information about how the product behaves outside the conditions available to the original design team.
The OEM can learn:
That information should feed back into engineering, product development, technical publications, parts planning and training.
The UK Ministry of Defence has long used the concept of a Data Recording, Analysis and Corrective Action System (DRACAS) to create this kind of closed loop. Its guidance describes how in-service events can be collected, analysed, corrected and tracked to improve products over time, lower whole-life cost and reduce operational risk.
The MOD reliability and maintainability guidance is archived and over 10-years old, but the principle remains highly relevant.
Downtime, availability and customer retention are important results, but they change after many operational decisions have already been made. OEMs also therefore need measures that reveal where the repair journey is slowing down.
Before comparing figures, you should define exactly when every clock starts, pauses and stops. MTTR (Mean Time To Repair) is particularly ambiguous because it can mean time to repair, restore, resolve or recover. Some definitions include detection and diagnosis, while others measure active repair only.
An OEM could use the following framework:
| KPI | Suggested calculation | What it reveals |
|---|---|---|
| Non-active repair time | Total restoration time minus active repair time | The delay surrounding the productive work |
| NART percentage | NART divided by total restoration time × 100 | How much of the customer’s downtime journey is non-active |
| Mean repair time (MTR) | Total active diagnosis, repair and test time divided by completed repairs | How efficiently the physical work is completed |
| Mean time to diagnose | Total active diagnostic time divided by diagnosed incidents | Whether technicians can isolate faults efficiently |
| Mean parts identification time | Time from parts search starting to correct part confirmation | Friction in catalogues, configuration and fitment data |
| Parts identification accuracy | Correct first selections divided by total part selections × 100 | How often the right part is identified first time |
| First-time fix rate | Repairs completed without repeat attendance or an additional parts order divided by total repairs × 100 | Whether diagnosis, parts and procedures combine into a successful outcome |
| Technical escalation rate | Incidents requiring OEM expert support divided by total incidents × 100 | How well knowledge scales beyond individual experts |
| Repeat repair rate | Repeat failures within an agreed period divided by completed repairs × 100 | Repair quality and unresolved root causes |
| Product knowledge update time | Time between an approved change and affected service content being updated | How quickly engineering truth reaches the front line |
| Feedback closure time | Time between a field issue being recorded and a decision or corrective action | Whether service outcomes are influencing future products |
Begin with one important product family or recurring repair type rather than trying to map the entire installed base at once. This is a much easier way to gain business-wide support for a more robust approach to improving access to product knowledge in the aftermarket.
Agree when the clock starts, what counts as active work, which pauses count as NART and when the product is genuinely back in operation.
Record the start and end of diagnosis, parts identification, approval, logistics, repair and testing. Avoid relying entirely on free-text notes.
Measure a representative period and calculate NART by stage. Compare product families and failure modes using consistent definitions.
Look for repeated searches, frequent escalations, wrong parts, missing procedures, stock questions and repeat visits. The largest delay may sit outside the physical repair.
Give technicians, dealers and support teams access to configuration-aware parts information and visual instructions at the point where they need to make a decision.
Capture what failed, what fixed it, which part was used, how long each stage took and where the guidance was unclear. Ensure that information is trustworthy, and then feed that information back into the product, the support content and the design process.
Track whether NART, MTR, escalations, mis-orders and repeat visits improve. The value comes from changing the journey, not from creating another dashboard.
When a machine fails, the customer does not separate active repair time from information searches, approval queues or parts delays - all they see is a product that they rely on is unavailable.
OEMs therefore need to better understand the stages inside that period.
Once non-active repair time is visible, teams can then identify which delays are necessary, which can be shortened and which should never have occurred. Digitised workflows to connect product knowledge to the right people is key as it helps them recognise the machine, find the right part, follow the correct process and capture what happened afterwards.
That shortens the immediate journey back to operation, and also gives engineering the evidence needed to make the next product easier to diagnose, support and repair.