MTTR
Mean time to repair (MTTR) is a maintenance metric that measures the average time required to troubleshoot, fix, and restore failed equipment to normal operating condition. Calculated by dividing total unplanned downtime by the number of failures, MTTR spans failure detection, notification, diagnosis, spare parts logistics, physical repair, and post-repair quality verification. Rather than isolating hands-on wrench time, MTTR captures the complete recovery process. Lower MTTR values reflect a more responsive maintenance system, directly increasing overall equipment availability and operational uptime.
- Detect: the clock starts
- This bar is one repair on the LT-3 leak tester, failure 2 of 4. The machine stopped at hour 330 of the month. The MTTR clock starts at that moment, not when a technician arrives. The first amber slab is detection: 0.5 hours before anyone knew the machine had stopped. Faster detection is often the cheapest MTTR gain.
- Diagnose: finding the cause
- The second amber slab is diagnosis: 1.5 hours to find a cracked seal cartridge. No repair happens yet. Diagnosis is the largest slab in this repair. Fault trees, sensor history, and a technician who knows the machine shrink it. Condition monitoring can shrink it further by naming the failing part before the stop.
- Wait for spare: not repair work
- The crimson slab is 1.0 hour waiting for a seal cartridge from the central store. Nobody touches the machine. It still counts in MTTR, because the machine is still down. Spare parts policy, not wrench skill, decides this slab. With the cartridge on the shelf beside the LT-3, the slab disappears.
- Verify: the clock stops
- The last indigo slab is verification. Test pieces run for 0.5 hours to confirm the tester passes good parts. MTTR ends here, when good parts run, not when the wrench is put down. Skipping verification shortens the number and lengthens the next failure. This repair totals 4.0 hours, or 3.0 hours with the spare on hand.
- The month average
- Mean time to repair averages over every repair in the month. Four repairs took 1.5, 4.0, 2.0, and 2.5 hours: 10 hours of downtime. MTTR is 10 divided by 4: 2.5 hours. With the spare on hand, this repair drops to 3.0 hours. Downtime falls to 9 hours and MTTR to 2.25 hours.
Key facts
- Calculation
- Total downtime ÷ number of failures
- Measurement start
- Moment of unplanned stoppage
- Measurement end
- Resumption of verified quality output
- Availability formula
- MTBF ÷ (MTBF + MTTR)
- Breakdown stages
- Detection, notification, diagnosis, logistics, repair, verification
By Matthew Savas — Founder of Kaizumi. Reviewed 1 September 2026.
Mean time to repair (MTTR) is a maintenance metric that measures the average time required to troubleshoot, fix, and restore failed equipment to normal operating condition. The measurement begins the moment an unplanned equipment stoppage occurs and concludes only when the asset resumes production and yields verified, quality output. MTTR encompasses every phase of the restoration process, including failure detection, technician notification, diagnostic analysis, procurement of replacement parts and tools, physical corrective action, and post-repair quality verification. Within manufacturing and asset management, lower MTTR values indicate a more responsive and efficient maintenance system, leading directly to higher equipment uptime.
Definition and mathematical calculation
MTTR quantifies maintainability by determining the average duration of unscheduled downtime events across a specific evaluation period. The calculation divides the total hours or minutes of downtime resulting from equipment breakdowns by the total count of failures observed during that same period.
To calculate MTTR, sum the total duration of all corrective maintenance interventions and divide that sum by the number of individual failure incidents. For example, if a machine stops three times in a week, resulting in individual downtime durations of two hours, four hours, and six hours, the total downtime is twelve hours. Dividing twelve hours by three failures yields an MTTR of four hours per event.
MTTR differs fundamentally from metrics that measure the physical duration of mechanical labor. In standard industrial practice, MTTR accounts for all organizational and logistical delays surrounding the breakdown event rather than isolating hands-on wrench time.
Stages of an equipment breakdown
An equipment stoppage consists of distinct consecutive segments. Analyzing downtime by individual stages allows maintenance and operations teams to isolate organizational delays from mechanical complexities.
- Detection time: The duration between the exact moment of machine failure and the moment personnel or automated monitoring systems identify the stoppage.
- Notification and response time: The interval required to alert the maintenance department and have a technician arrive at the equipment.
- Diagnostic time: The period spent inspecting the machine, reading fault logs, conducting electrical or mechanical checks, and pinpointing the root cause of the failure.
- Logistical and administrative delay: The time spent waiting for replacement parts from stores, obtaining specialized tools, completing work permits, or waiting for external technical support.
- Physical repair time: The actual hands-on mechanical, electrical, or software work performed by the technician to replace damaged components or correct settings.
- Verification and ramp-up time: The time required to test the machine, run calibration samples, inspect test pieces, and verify that the machine produces parts according to quality specifications before handing it back to production.
In many industrial breakdowns, the physical repair time represents only a small portion of the total downtime window.
Worked example on an assembly line
To illustrate the composition of MTTR, consider the LT-3 leak tester operating on an automotive valve assembly line over a standard 720-hour operating month. During this month, the LT-3 experienced four separate unplanned breakdowns.
The second failure of the month occurred at hour 330 and generated a total downtime of 4.0 hours. A breakdown of this specific event shows the following individual time intervals:
- Detection took 0.5 hours because the machine stopped during a night shift and operators did not immediately notice the stoppage.
- Diagnosis took 1.5 hours while the technician systematically inspected the pneumatic circuit and traced the fault to a cracked seal cartridge.
- Waiting for spare parts took 1.0 hour because the replacement cartridge was stored in a central warehouse on the opposite side of the plant.
- Physical repair took 0.5 hours to remove the damaged cartridge, clean the seating surface, and install the new seal.
- Verification took 0.5 hours to run test pieces and confirm that the machine accurately measured leak rates within specification.
In this 4.0-hour breakdown event, hands-on mechanical work accounted for only 0.5 hours, while diagnostic, logistical, and administrative steps accounted for the remaining 3.5 hours.
Across the entire 720-hour month, the four breakdown events on the LT-3 leak tester lasted 1.5 hours, 4.0 hours, 2.0 hours, and 2.5 hours. Summing these values gives a total downtime of 10.0 hours.
To calculate the mean time between failures (MTBF), subtract the 10 hours of downtime from the 720 total available hours to obtain 710 operating hours, then divide by the 4 failures to reach an MTBF of 177.5 hours. Availability is calculated by dividing the 710 operating hours by the 720 total planned hours, resulting in 98.6 percent.
Now suppose the plant implements a lean improvement by storing a designated spare seal cartridge in a point-of-use cabinet directly beside the LT-3 tester. During an identical failure scenario, the 1.0-hour waiting delay for parts disappears. The total duration of repair 2 drops from 4.0 hours to 3.0 hours. Total monthly downtime drops from 10.0 hours to 9.0 hours, reducing the monthly MTTR from 2.5 hours to 2.25 hours, while the machine MTBF remains unchanged at 177.5 hours.
Relationship to MTBF, availability, and OEE
MTTR operates alongside MTBF to determine the theoretical and practical availability of production equipment. While MTBF measures the reliability of an asset by tracking how long it runs between successive breakdowns, MTTR measures the speed of recovery when an asset stops.
The mathematical relationship between MTBF and MTTR directly defines equipment availability:
To find availability using these reliability metrics, divide MTBF by the sum of MTBF and MTTR. When MTTR decreases while MTBF remains constant, overall availability increases.
In Total Productive Maintenance (TPM), availability serves as one of the three foundational components of Overall Equipment Effectiveness (OEE), alongside performance and quality. Practitioners analyzing OEE, the metric TPM moves, focus on MTTR reduction as the primary method to eliminate availability losses caused by unplanned equipment stops. Teams tracking performance across lines can verify these operational relationships directly using an OEE calculator.
Significance in lean operations
Tracking MTTR provides organizations with insight into the broader operational infrastructure rather than just individual technician performance. A breakdown that lasts 4.0 hours with only 0.5 hours of hands-on mechanical activity reflects systemic inefficiencies in detection mechanisms, diagnostic resources, and spare parts logistics.
MTTR functions as the fastest lever for recovering lost production capacity. Improving MTBF often requires engineering redesigns, component upgrades, or extensive lubrication and maintenance overhauls that take months to implement. In contrast, reducing MTTR can be achieved quickly by reorganizing maintenance support workflows, improving spare parts positioning, and standardizing troubleshooting documentation without altering machine hardware.
Methods for reducing MTTR
Organizations implement specific technical and organizational standards to systematically decrease the individual components of repair time:
- Automated detection and visual management: Installing andon systems, automated line-monitoring sensors, and stack lights ensures immediate notification of stoppages, cutting detection latency.
- Standardized diagnostic trees: Developing visual troubleshooting guides and step-by-step diagnostic flowcharts for common failure codes allows technicians to isolate root causes without extensive exploratory testing.
- Point-of-use spare parts storage: Relocating critical replacement components, fasteners, and specialized tooling to shadow boards or storage cabinets adjacent to the machine eliminates transit time to central stores.
- Quick-change mechanical design: Utilizing captive fasteners, quick-release pneumatic couplings, and modular sub-assemblies allows technicians to replace entire assemblies rapidly and rebuild damaged units offline.
- Standardized testing procedures: Creating precise checklists for post-repair verification ensures that startup scrap is minimized and production resumes immediately after physical work finishes.
Industry applications
Manufacturing
A cross-functional maintenance team on a high-speed packaging line analyzed equipment breakdown records and identified an average MTTR of 4 hours across all packing stations. A time-study breakdown revealed:
- 30 minutes to detect failures during high-noise production intervals
- 45 minutes for off-site maintenance personnel to respond to calls
- 90 minutes of electrical and mechanical fault diagnosis
- 60 minutes of physical replacement work
- 15 minutes of test running and quality sign-off
The facility restructured its staffing model to implement dedicated line-side maintenance coverage and installed visual andon displays. These changes eliminated transit delays and reduced response times, lowering the overall MTTR from 4.0 hours to 2.5 hours.
Healthcare
A biomedical engineering department evaluated medical diagnostic equipment downtime and found that repairs averaged 6 hours per incident. Detailed process mapping revealed that 2 hours of every repair event were spent requisitioning, picking, and transporting electronic components and replacement cables from a centralized basement storeroom.
By establishing point-of-use spare part stations within clinical department service bays, the hospital reduced part retrieval time to 20 minutes, driving overall biomedical MTTR down from 6 hours to 4 hours.
Administrative and information systems
An enterprise IT department measured an average service restoration time of 90 minutes following critical database server interruptions. Process tracing showed the recovery time consisted of:
- 10 minutes for user reports to establish an outage
- 15 minutes to assign an on-call systems engineer
- 30 minutes to review log files and diagnose configuration or database errors
- 25 minutes to execute corrective scripts and restart services
- 10 minutes to verify data integrity and service responsiveness
The organization deployed automated infrastructure monitoring to detect outages within 2 minutes and published standardized troubleshooting runbooks that reduced diagnostic time to 15 minutes, lowering total MTTR to 65 minutes.
Common measurement errors
Organizations tracking MTTR often encounter systematic measurement errors that distort performance data:
- Excluding diagnostic and logistics time: Recording only the hands-on mechanical wrench time instead of the full stoppage-to-production interval artificially lowers reported MTTR and conceals logistical waste.
- Omitting quality verification: Halting the repair timer the moment physical reassembly finishes rather than after successful verification of first-piece quality leads to premature sign-offs and frequent secondary stops.
- Tracking aggregate averages without distribution analysis: Relying strictly on a single plant-wide average can mask extreme outliers; several minor 10-minute stops can statistically hide a recurring 8-hour catastrophic failure.
- Combining planned maintenance with breakdowns: Conflating scheduled preventive maintenance tasks with corrective breakdown interventions invalidates MTTR data, as planned maintenance does not measure unscheduled repair capability.
Frequently asked questions
- Does the MTTR clock stop when a technician finishes the physical repair?
- No, MTTR tracking continues until post-repair testing and quality verification are complete. The metric measures the entire duration until the asset resumes production and consistently yields verified, quality output. Halting the timer immediately after mechanical reassembly conceals startup scrap and risks secondary breakdowns.
- How does MTTR relate mathematically to equipment availability and MTBF?
- Equipment availability is calculated by dividing mean time between failures (MTBF) by the sum of MTBF and MTTR. While MTBF tracks how long equipment operates between successive breakdowns, MTTR measures how quickly operations recover when an asset stops. Consequently, reducing MTTR directly increases operational availability even when failure frequency remains unchanged.
- Why is reducing MTTR generally faster than increasing MTBF?
- Improving MTBF typically requires engineering redesigns, component upgrades, or extended maintenance overhauls that take months to implement. In contrast, reducing MTTR focuses on streamlining diagnostic workflows, standardizing procedures, and relocating spare parts to point-of-use storage. These organizational adjustments recover lost production capacity quickly without altering machine hardware.
- Should scheduled preventive maintenance be included in MTTR calculations?
- No, planned preventive maintenance tasks must be excluded from MTTR calculations. MTTR is strictly designed to evaluate response and restoration performance during unscheduled equipment breakdowns. Conflating scheduled servicing with unplanned corrective interventions distorts the data and conceals true breakdown recovery capabilities.
- Why is measuring only hands-on wrench time considered an error in MTTR tracking?
- Physical repair frequently accounts for only a small portion of total downtime, while detection, fault diagnosis, and waiting for spare parts consume the rest. Recording only hands-on wrench time artificially lowers reported repair metrics and hides administrative and logistical waste. Tracking the full stoppage-to-production interval ensures systemic operational bottlenecks are exposed and addressed.