Over 12 months of using eZOP, record accuracy rose from 62% to 99.4%, the number of overdue assets fell by three quarters, and the investment paid off on average in less than nine months.
Fire protection regulations specify precisely what must be inspected, how often and by whom. The problem lies in practice: data is kept on paper cards, in binders, spreadsheets and service providers’ reports. Records like these do not warn when a deadline is approaching, do not reliably preserve an asset’s history and do not give a complete picture of the system’s condition. The consequence is not just an administrative burden, but also a greater likelihood that an extinguisher will not work when a fire breaks out.
That is why we conducted a study measuring what actually changes when records of fire protection equipment and installations are moved into eZOP.
Why paper records don’t work
The Serbian Law on Fire Protection requires organizations to have their fire extinguishing and fire alarm installations regularly inspected by an authorized legal entity (as a rule, every six months), to keep records and to present them to inspectors. Fines for legal entities range from RSD 300,000 to RSD 3,000,000, and poorly kept documentation can also become grounds for disputing an insurance claim.
In practice, data on a single extinguisher lives in three places: on the label, in the service provider’s report and in the spreadsheet of the person responsible for fire protection (LZZP). There is no link between them, so everything depends on the diligence of one person. Before the system was introduced, we recorded:
- 7.9% of assets whose recorded location did not match the actual one;
- 214 cases in which the inspection report could not be found (and no report means no inspection);
- 18.4% of assets with an overdue inspection or service;
- unknown maintenance cost per asset;
- knowledge of the system’s condition concentrated in a single person;
- in multi-site companies, management without a consolidated overview.
What eZOP changes
In eZOP, the basic unit is not a document but an asset. Every extinguisher, hydrant, fire alarm control panel or emergency light gets a barcode, a unique identifier and its own history. Because the deadline belongs to the asset, the system calculates the next inspection itself and sends alerts 60, 30 and 7 days before it is due. Defects are logged with a photo, a deadline and a responsible person, and tracked until closed. The documentation package for inspectors is generated automatically, and every change is logged with the user’s name and a timestamp. The mobile app also works offline, in halls, basements and warehouses.
How we measured
We monitored 48 facilities across five sectors (industry, retail, logistics, office buildings and public facilities) with a total of 20,002 pieces of equipment. Fire extinguishers make up the largest share (59%), followed by emergency lighting, hydrants, electrical and lightning protection installations, and fixed fire detection and suppression systems. The same facilities were monitored for 12 months before and 12 months after the introduction of eZOP, across eight indicators.
Key results
All eight indicators improved:
- Record accuracy: 62.0% → 99.4%
- Overdue assets: 18.4% → 4.3% (−76.6%)
- Equipment passing inspection: 91.2% → 97.8%
- Service provider response to a reported defect: 9.6 → 3.1 days (−67.7%)
- Inspection preparation: 31.5 → 15.4 hours per facility (-51.1%)
- LZZP administrative work: 34.0 → 12.6 hours per month (−62.9%)
- Equipment location accuracy: 92.1% → 99.4%
- Annual maintenance cost per asset: RSD 2,480 → RSD 2,115 (−14.7%)
- Record accuracy rises fastest in the first three months, when the inventory uncovers equipment that was unrecorded or recorded in the wrong location. From the ninth month on, it stays above 97% without any extra effort.
The time saved comes from searching for reports, copying data and compiling reports by hand – tasks that have practically disappeared. Time spent on facility walkthroughs and working with employees, the part of the job that directly reduces risk, increased by 3.8 hours per month. Savings are greatest in retail chains (31.2 hours per month) and smallest in office buildings (15.3 hours).
Return on investment
We calculated conservatively: only working time saved, fewer emergency call-outs, and more efficient purchasing and refilling. Avoided fines and fire damage, which represent the greatest potential benefit, were not included.
For an average facility, the net benefit is around RSD 31,000 per month, and the investment pays off in 8.6 months (ranging from 6.1 to 11.4 months). The more complex the system, the faster the payback.
From reactive to predictive maintenance
After one year, the database contained more than 41,000 actions performed on equipment, revealing patterns that hold across all sectors:
- extinguishers in damp areas and areas with temperature fluctuations (boiler rooms, laundries, loading ramps) lose pressure 2.3 times more often;
- extinguishers along forklift routes are damaged or moved several times more often;
- an asset with a previous defect is 1.8 times more likely to have the defect recur;
- where servicing is delayed by more than seven days, equipment condition is systematically worse.
- This means inspection intervals can be tightened where the risk is higher, while keeping the statutory minimum everywhere else.
The number of recorded defects rose from 3,146 to 4,912, but this is not a deterioration: the system now captures what previously went unnoticed. More importantly, the share of defects left unresolved for more than 30 days fell from 41.2% to 6.8%, and the share of those caused solely by an expired deadline fell from 38.4% to 9.2%.
On a risk matrix of seven typical scenarios (an out-of-date extinguisher at the moment of a fire, a hydrant network without pressure, faulty emergency lighting, documentation unavailable during an inspection, and others), the aggregate exposure index fell from 99 to 41 points (-58.6%). Software cannot prevent a failure, but it significantly reduces the likelihood that a failure goes unnoticed until the moment the device is needed.
Three real-world examples
Food industry production plant (1,180 assets). The inventory uncovered 74 unrecorded extinguishers and 38 in the wrong locations. Overdue deadlines fell from 23.4% to 3.1%, and inspection preparation from 42 to 16 hours. In the boiler room, where the data showed more frequent pressure loss, the inspection interval was shortened to three months and six unsuitable extinguishers were replaced. Payback: 7.2 months.
Logistics center (46,000 m², 940 assets). The service provider previously responded in 11.2 days; with work orders and automatic notifications, this dropped to 2.8 days. Emergency call-out costs fell by 31%. Response times shortened as soon as they became measurable, even before any contractual measures. Payback: 9.4 months.
Retail chain (6 locations, 2,410 assets). Each store kept its own records, and compiling a report for the chain took weeks. Record accuracy rose from 55% to 99.1%, head office saves 31.2 hours per month, and every location can prepare inspection documentation within 24 hours. Payback: 6.1 months, the shortest in the sample.
Recommendations
For companies:
Adopt digital records as standard, regardless of the facility’s fire risk category. Before going live, inventory and barcode-tag every asset. This is the single most important success factor: facilities that devoted enough time to the inventory reached their targets two months earlier on average.
Define in an internal policy who responds to automatic alerts and within what time frame. Contractually require the service provider to enter findings into the system within 48 hours. Track five indicators monthly: record accuracy (≥ 99%), overdue deadlines (≤ 2%), defects older than 30 days (0), service provider response time (≤ 5 days) and inspection readiness (≤ 24 hours).
For multiple locations, introduce a consolidated overview as the basis for the maintenance budget. For regulators: recognize digital records with a logged change history as an equally valid form of keeping mandatory records, standardize the electronic inspection report, encourage digitalization in facilities of the first and second fire risk categories, and establish a national database of anonymized failure data.
What’s next
The study also pointed to directions for eZOP’s development: IoT sensors for continuous pressure monitoring in extinguishers and hydrant networks, ranking assets by likelihood of failure, verifying that an asset is in place between two inspections, standardized export of inspection documentation, integration with CMMS/BMS systems, tracking employee training, and benchmarking against sector averages.
Conclusion
Moving from paper to digital records is not just a modernization of administration. Records stop being an after-the-fact log of work done and become a mechanism that triggers the work itself before the deadline expires. Meeting obligations no longer depends on someone remembering in time.
Note: The results are based on data from pilot implementations and benchmark values from the practice of authorized service providers, and will be further validated on an expanded sample.
