Structured maintenance linked to 99.91% rolling-door availability in Lima

Ten industrial rolling doors at a manufacturing plant in Lima, Peru, recorded substantially higher availability after the plant introduced a structured maintenance-management program centered on quarterly preventive maintenance. In an uncontrolled before-and-after comparison of two six-month periods, mean operational availability rose from 85.21% to 99.91%. Mean time between failures, or MTBF, increased from 359.61 to 2,338.45 hours, while mean time to repair, or MTTR, fell from 48.89 to 1.55 hours. Recorded failures declined from 137 to 25, and total downtime fell from 6,390 to 38 hours. All ten doors changed in the same direction on the three principal indicators, but the study design cannot establish that the maintenance program alone caused the improvement.

The research was designed as an uncontrolled longitudinal case study in which each rolling door served as its own comparison. The baseline covered January through June 2024, and the post-implementation period covered July through December 2024. Program preparation began on 1 May 2024, continued through May and June as a two-month run-in, and was considered fully implemented on 1 July. The researchers included assets in routine service only when records for scheduled time, operating time, downtime and failures were complete in both periods. All 10 selected doors had complete paired data, so no statistical imputation was used. Each door contributed 4,320 nominal scheduled hours per period, equivalent to six months of calendar-based service exposure. Because there was no concurrent control group and no random assignment, the results establish a temporal association rather than definitive causal evidence.

Industrial rolling doors are auxiliary assets rather than primary production machines, yet their mechanical, electrical, control and safety systems affect whether personnel, vehicles and materials can move through a plant. A failed door can interrupt material flow, restrict access and, in some configurations, create safety concerns. The researchers noted that empirical maintenance evidence for industrial access systems remains limited, while much of the surrounding reliability literature concentrates on maintenance optimization, condition monitoring, predictive maintenance and primary production equipment. This case therefore examined a less-studied class of support equipment using routine corrective-maintenance reports, operational logs, direct inspections and paired asset-level records rather than continuous sensor data or a predictive-maintenance platform.

The intervention began with a diagnostic review of historical reports and technical inspections. Maintenance practitioners considered mechanical, electrical, control and safety components, then organized possible contributors to downtime using an Ishikawa diagram. A judgment-based pairwise matrix ranked the absence of a preventive-maintenance plan highest, with 30 of 139 prioritization points. Lack of inspection control followed with 20 points, components in poor condition with 18, high technical-staff turnover with 16, and lack of appropriate instruments with 13. These were engineering-prioritization scores, not measured percentages of failures. The assessment was conducted by a plant maintenance mechanic and a maintenance supervisor, but the source records do not establish whether they rated independently, how disagreements were resolved or whether agreement was formally quantified. Asset criticality was also assessed through ordinal judgments based on failure frequency, operational impact, operational flexibility, relative maintenance cost and safety, health and environmental consequences. Those classifications organized maintenance priorities and were not externally validated risk probabilities.

Functional failure analysis documented problems involving brakes, sensors, anti-impact safety edges, controllers, variable-frequency drives, manual operating mechanisms, guides and bearings. Depending on the door and component, recorded failure modes could immobilize the door, prevent manual operation, cause sudden movement, generate errors or create crushing or impact hazards. The resulting quarterly maintenance program specified the activity, required door position, estimated duration, frequency and responsible technician. Tasks included cleaning and inspecting motors and electrical cabinets, tightening terminals and power and control connections, measuring voltage and current, adjusting encoders, opening and closing limits and motor brakes, inspecting manual systems, lubricating bearings and mechanical components, and functionally checking vehicle-detection sensors, magnetic sensors, photocells and anti-impact safety edges. Maintenance work was performed with the equipment de-energized under the plant’s lockout and industrial-safety procedures.

The program was broader than the execution of preventive tasks alone. It also included four training sessions involving four workers—two sessions on maintenance tasks and two on equipment—as well as standardized record keeping and closer operational follow-up. The study consequently could not separate the effects of lubrication, adjustment, cleaning, training, documentation, monitoring and any corrective component replacements made during follow-up. A complete door-by-door compliance log was unavailable, preventing the researchers from quantifying adherence, identifying missed or delayed tasks or establishing what proportion of scheduled work was completed. This incomplete fidelity trail is one reason the results should be interpreted as an association with a comprehensive maintenance-management program rather than a controlled estimate of the effect of a single procedure.

The reliability calculations used calendar-based service availability. Every month was standardized to 720 hours, and service-available time was calculated as nominal scheduled time minus downtime. Downtime was measured from notification of a failure or operational stoppage until restoration to service. Preventive activities performed during planned maintenance windows were documented separately and were not counted as unplanned failures. Because opening and closing cycles, forklift traffic, pedestrian traffic, active motor runtime and other workload measures were unavailable, MTBF could not be interpreted as a cycle-normalized component lifetime. Instead, it was a reliability indicator based on nominal calendar exposure. The fixed denominator made the windows comparable for the studied doors, but it also meant that utilization differences could not be distinguished from calendar-based availability.

All 10 doors had higher availability and longer MTBF during the second period, accompanied by shorter MTTR. The mean paired increase in availability was 14.70 percentage points, with a 95% bootstrap confidence interval of 10.83 to 18.75 percentage points. Mean time between failures increased by 1,978.84 hours, with a 95% confidence interval of 1,199.74 to 2,868.94 hours. Mean time to repair decreased by 47.34 hours, with a 95% confidence interval from a 57.05-hour reduction to a 39.92-hour reduction. Exact two-sided paired Wilcoxon signed-rank tests produced p = 0.002 for each of the three principal comparisons. Matched-pairs rank-biserial correlations were +1.00 for availability and MTBF and −1.00 for MTTR, showing that every paired change had the same direction within this sample. Those effect-size descriptors do not establish that the same magnitude would be observed at other plants.

Several limitations constrain how far the improvement can be generalized. The study covered only 10 assets at one manufacturing plant, used two six-month windows and had no concurrent comparison group. Seasonal changes, traffic variation, equipment replacement, changes in organizational practice and other concurrent developments therefore cannot be separated from the maintenance program. The January-to-June baseline was not completely untreated because implementation preparation and a two-month run-in overlapped its final two months, potentially reducing the measured contrast. Post-implementation availability was near its upper limit, and the short observation horizon limits inference about long-term reliability. Other limitations include the missing door-by-door adherence log, the inability to reconstruct all corrective work, the absence of usage-intensity data and the use of calendar rather than cycle-based exposure. Replication with longer follow-up, electronic work orders documenting task completion, recorded opening and closing cycles, traffic intensity and concurrent comparison assets would be needed to test whether the pattern persists and to distinguish the program from other changes.

For maintenance organizations without advanced condition-monitoring systems, the case shows how routine records can support a quantitative asset-level baseline when definitions and exposure periods are explicit. The sequence used by the plant—diagnosis, cause prioritization, criticality assessment, failure-mode analysis, task design, training and standardized record keeping—converted scattered repair reports into paired measures of availability, failure frequency and repair duration. The most striking operational contrast was between 6,390 hours of downtime before full implementation and 38 hours afterward, accompanied by a reduction from 137 recorded failures to 25. Those figures document a large improvement for the ten doors in this plant, but they do not constitute a universal forecast or a controlled estimate of the program’s causal effect. The defensible conclusion is that introducing an organized preventive-maintenance program was associated with large and consistent improvements in the recorded performance of these industrial rolling doors during the following six months.

Subject of Research: Paired before-and-after study of preventive maintenance and operational availability in 10 industrial rolling doors at one Lima manufacturing plant.

Article Title: Changes in operational availability following a structured preventive-maintenance program for industrial rolling doors: a before-and-after case study in lima, Peru

Article References: Calero Obregón, R. C., Cuba Vargas, K. E., Gamarra Tolentino, J. E., Mendoza Zenozain, E. A., & Corpus Vergara, N. B. (2026). Changes in operational availability following a structured preventive-maintenance program for industrial rolling doors: a before-and-after case study in lima, Peru. Frontiers in Mechanical Engineering, 12, Article 1991577. https://doi.org/10.3389/fmech.2026.1991577

Image Credits: AI Generated

DOI: 10.3389/fmech.2026.1991577

Keywords: preventive maintenance, industrial rolling doors, operational availability, MTBF, MTTR, reliability engineering, maintenance records, asset management, factory downtime, Peru, Changes, operational