Most devices don't wear out. The environment gets them first.

The failure data doesn't match the engineering story

Electronic devices are better engineered than they have ever been: more reliable silicon, batteries that hold more charge for longer, thermal design that has matured into a discipline of its own. Manufacturing tolerances that were aspirational fifteen years ago are now routine on high-volume lines.

Field failure rates have not fallen at the same pace.

That gap has an explanation, and it is not build quality. Most devices no longer fail because a component wore out. They fail because of what surrounded them: humidity in a coat pocket, condensation in a chilled warehouse aisle, sweat against a hearing aid worn through a workout, salt air around a sensor bolted to a harbour crane. Liquid damage remains the single most common cause of electronics field failure, and it accounts for a significant share of warranty returns, replacement costs and product recalls in every major electronics sector.

The industry's engineering effort has concentrated on the device. The failure data points somewhere else entirely: at the environment the device has to live in.

Four environments, four different failures

Consider what "the environment" actually means across four categories.

A hearing aid spends roughly 16 hours a day against human skin, the warmest and most humid microclimate any consumer electronic device routinely operates in. Sweat, ear canal moisture, handling oils, and condensation every time the wearer moves between cold outdoor air and a heated room. Standard reliability testing (temperature cycles, humidity chambers, drop tests) does not simulate any of it. One manufacturer we worked with was seeing corrosion at multiple points inside devices that already carried a conformal coating. When protection was redesigned around the environment the device actually faced, field failure returns fell by half.

A vehicle infotainment unit lives with cabin condensation that pools directly on its circuit boards. A conventional 50-micron acrylic spray cannot reach underneath densely packed components or coat both sides of a board adequately, which is why corrosion keeps appearing on boards that were nominally protected. Getting this right at automotive scale means passing IEC 60068 temperature and humidity cycling, and holding that standard across 1.5 million devices a year.

An industrial sensor estate is a different problem again. A single chemical plant can run 5,000 sensors across 200 distinct microenvironments: vibration, chemical exposure, condensation, particulates, temperature swings. The maintenance team that used to service 50 fixed-location PLCs now covers 5,000 endpoints it can barely visit. Once installation and replacement labour are counted, a field failure in industrial electronics can cost up to 100 times the original assembly cost.

An electronic shelf label spends its life in retail's chilled aisles: damp air, daily condensation cycles, spray-wipe cleaning chemicals. The result is creep corrosion and dendrite growth, the mechanisms behind the intermittent faults that make a fleet of thousands of labels expensive to run. Gaskets alone don't stop moisture reaching the electronics, and potting protects the board at the price of making repair impossible.

Same physics in every case. Very different manifestations, and no single protection answer that covers them all.

Protection arrives too late

Here is the structural problem: in most device programmes, protection is a late-stage question. It surfaces as a qualification conversation, often three months before production sign-off. By then the device architecture is locked, the manufacturing line is specced, and the realistic options have narrowed to whatever fits the decisions already made.

A coating chosen at that point is a compromise by definition. It has to work around the enclosure, the component layout, the line takt time and the test plan, none of which were designed with it in mind. The device gets whatever protection is still possible, rather than the protection its environment demands.

The reframe: protection as a platform decision

If the environment is the failure agent, the protection question changes shape. It stops being "which coating do we apply at the end?" and becomes "what does this device need to survive, and what does that imply for how it is designed and built?"

Answering that question honestly requires more than one technology. Plasma coating deposits protection at the molecular level, penetrating geometries that dispensed liquids and mechanical seals cannot reach, with batch precision suited to fully constructed devices. Liquid barrier coating integrates directly into a production line at speed. And some environments are new enough that the right answer is a process developed specifically for that device and that line. The starting point is never a product name. It is the qualification test plan and the failure modes the device will actually face, mapped against every available protection path. In some applications, the answer is two technologies deployed together.

Treated this way, protection stops being a cost line and starts shaping the design itself. The coating choice changes the available form factor: a device protected from the inside can carry openings, hinges and vents that a seal-dependent design cannot. The form factor changes the thermal envelope. The thermal envelope changes the component choices. Engineers who bring protection into the conversation at specification, rather than qualification, find the rest of the design space opens up.

This is the argument for a platform rather than a product, and it is the position P2i has built: the only engineering partner offering both plasma and liquid protection technologies at OEM scale, globally, with the engineering process to determine which is right for each application.

What the record shows

The approach is not a thesis. Over twenty years, P2i has protected more than ten billion devices, currently adding over 100 million a year, at a 99.99% first pass yield in production. Numbers at that scale prove a specific point. They demonstrate that environment-first protection survives contact with real production: real yield requirements, real line speeds, and qualification regimes that do not bend, across environments as different as an ear canal and a chemical plant.

Where to start

The practical shift is timing. Protection belongs in the specification conversation, alongside form factor and thermal design, not in the qualification scramble at the end. That costs one conversation with an application engineer early in the programme. The alternative (a locked architecture that cannot accommodate the protection its environment demands) costs field failures for the life of the product.

Devices are not going to gentler places. Protection design should get there first.