Repair, don't replace: managing the ATM life cycle in banking

By the Redsis Technical Services Division team · October 2026 · 6 min read

A channel that's shrinking, not disappearing

According to the 2025 Febraban Banking Technology Survey, Brazilian banks processed 208.2 billion transactions in 2024, of which 3.6 billion took place at ATMs. That's just 1.7% of the total and 13% fewer than the year before. Withdrawals fell 18% and deposits 22%.

Read in isolation, those numbers might suggest the ATM is on its way out. But 3.6 billion transactions a year is still close to ten million a day. And the people who depend on cash — small merchants, rural communities, customers with limited digital access — have no immediate substitute.

For a bank, that creates a demanding equation: keep availability high across a fleet that isn't growing the way it used to, without driving up cost per unit. That's where maintenance strategy makes the difference.

The hidden cost of "swap the whole module"

Under many traditional contracts, when a module fails it's replaced outright and the faulty part is scrapped or shipped back to the manufacturer. It's fast, but it creates three problems:

• Cost. A new module can cost several times more than repairing the component that actually failed.
• Parts availability. For older generations, new modules can be scarce or come with long lead times.
• Waste. Electronics that could run for years end up in the scrap bin.

The alternative is component-level repair: diagnose the real fault, fix the board, power supply or mechanism, and return it to the fleet after manufacturer-equivalent testing. It takes a lab, engineering and quality control — but it completely changes life-cycle economics.

What gets repaired inside a modern ATM

Today's ATM has little in common with the one from twenty years ago. For our client, repair scope spans very different equipment families:

• Cash recyclers and dispensers. Recyclers reuse deposited cash to fund withdrawals, cutting replenishment trips but adding mechanical and note-validation complexity.
• Peripherals such as card readers, printers and keypads.
• Power supplies and motherboards — the electronic core of the machine.
• Touchscreen monitors , the customer interface and one of the most wear-prone parts.
• AI-enabled cameras that add monitoring and security capabilities.
• Electronic locks for safes and cash modules.
• Fog generators — systems that protect cash during physical attack attempts.

Each family has its own technology, failure modes and testing requirements. That's why a banking repair partner needs more than technicians — it needs engineering capability.

Five lessons for managing an ATM fleet's life cycle

1. Treat repair as engineering, not a workshop

When a bank adopts a new technology, there's rarely a documented repair procedure in the market. For this client, our engineering team develops new technologies and builds repair processes for next-generation equipment. That keeps new components from becoming effectively "unsupported" once the warranty ends.

2. Lock in spare parts before you need them

Fleet availability depends on parts as much as on labor. A contract that combines repair and spare-parts supply under one accountable partner cuts downtime and ends the debate over who was supposed to have the part in stock.

3. Hold repairs to manufacturer standards

A repaired component is only worth it if it performs like a new one. Testing, traceability and quality control should match the original manufacturer's. At this client, that standard is formalized as a status: Redsis holds the same standing as the OEMs within the bank.

4. Keep physical security in scope

Cameras, locks and fog generators aren't accessories — they protect cash and people. Leaving them out of the maintenance plan creates a weak spot exactly where the risk is highest.

5. Measure by life cycle, not by repair ticket

The right metric isn't what it costs to fix one part; it's what it costs to keep a unit available over its entire service life. Component-level repair, refurbishment and reuse typically win that comparison against a replacement-only model.

The outcome: a manufacturer-level partner

Today Redsis serves this bank under an exclusive repair and spare-parts contract, with the same standing as the original manufacturers. Our engineers build repair processes for next-generation equipment, and our scope runs from cash recyclers and dispensers to AI cameras, electronic locks, fog generators, motherboards and touchscreen monitors.

For a bank that serves millions of customers, that means something concrete: one accountable technical partner keeping the fleet alive, with the engineering depth to keep up with each new generation of technology.

Where to start

If your institution swaps entire modules for faults that could be repaired, struggles with long lead times on spare parts, or has new equipment with no clear post-warranty support path, it's worth revisiting the model. A good first step is a failure analysis of the past year by equipment family — it usually shows where the biggest repair opportunity lies. At Redsis, we pair more than 25 years of experience with Latin American banks with a Technical Services Division dedicated to self-service hardware.