Ask an operations team how many pressure relief devices lifted on their plant last month, and in most facilities nobody will be able to tell you. A plant can have a full historian, a digital twin and thousands of instrumented tags, and its relief devices will still say nothing: no lift, no timestamp, no duration, no quantity released. It’s the last blind spot on an otherwise fully instrumented site, and it’s a big part of why risk-based inspection still runs on assumption when it could be running on evidence.

Inspection in asset-intensive industries has always been about managing uncertainty. Operators need to know not just whether an asset might fail, but what that failure would cost in safety, production and money. Over the last twenty years, Risk-Based Inspection (RBI) has become the standard way to handle that: a structured way of prioritising inspection effort according to risk rather than the calendar.

Combine probability of failure with consequence of failure and you get a method for putting attention where it actually matters, instead of inspecting everything on a fixed schedule regardless of condition. That shift results in lower inspection spend, better reliability, stronger safety performance.

The limitations of RBI

But RBI has a built-in limitation as it’s built on snapshots with the condition assessed at discrete points in time, with uncertainty creeping back in during the gaps. Risk profiles get updated periodically, rather than continuously and degradation can develop quietly between assessments. An asset rated low-risk at the last inspection might already be behaving differently by the time the next one comes round. Even a well-run RBI programme is, underneath it, a model working with incomplete data.

For a lot of operators that creates an ongoing tension. There’s pressure to cut inspection costs and stretch intervals wherever possible, but also a need to stay confident that critical assets are doing what they’re supposed to. The usual compromise: over-inspect low-risk kit for reassurance and accept some exposure to failures that happen between intervals anyway.

That tension has a real cost. Manual inspection programmes are labour-intensive and disruptive, especially on large or complex sites, and they’re getting harder to staff at any price: the Global Energy Talent Index 2026 puts 48% of the energy workforce at 45 or over, with 28% of lead operators already 55-plus. The constraint is shifting from what manual rounds cost to whether there’s anyone left to walk them. Meanwhile unplanned failures cost far more than the inspections that might have caught them – not just in lost production, but in safety, environmental impact and regulatory attention. The real problem isn’t cost versus risk, it’s that nobody can see what’s actually happening between inspections.

That gap is where the next stage of inspection is heading. Instead of tweaking inspection intervals or refining the modelling, more operators are asking a blunter question; what if you could observe asset condition continuously, instead of inferring it periodically?

The move towards continuous monitoring

Continuous monitoring moves things from static assessment to dynamic awareness. Rather than relying on historical data and predictive models alone, operators get a live read on how assets are actually behaving. Risk stops being something recalculated at the next inspection cycle and becomes something you can watch, check and act on as it happens.

Two asset types make this especially clear: pressure relief devices (PRDs) and slow-rotating bearings. Both are hard to monitor, for different reasons. PRDs are silent by design, only acting during abnormal conditions, and a conventional device gives no feedback at all when it does act. After the event, most operators can’t say with any confidence whether a valve lifted, at what pressure, for how long, or how much was released. Slow-speed bearings have the opposite issue, they can run for a long time while damage builds up and conventional vibration monitoring often isn’t sensitive enough to catch it early. Either way, operators end up leaning on periodic inspections, manual checks or plain assumption about condition, which leaves a gap between inspections and pushes maintenance strategy towards the conservative end by default.

The effect of increased regulation

Regulation isn’t moving in one direction globally, but where it does bite, it bites hard. In the EU, Regulation (EU) 2024/1787 has been in force since August 2024 and requires source- and site-level methane quantification, verified by an independent third party, plus leak detection and repair obligations, with penalties that can reach 20% of annual turnover.

In the US, the Petroleum Refinery NESHAP goes further specifically for relief devices: 40 CFR 63.648(j) requires each covered device to carry a system that identifies a release, records its time and duration, and notifies operators immediately, and it treats a second release from the same device for the same root cause within three years as a violation on its own. Across much of the Gulf and Asia, the pressure is commercial rather than legal, driven by OGMP 2.0 Level 4 and 5 source-level measurement rather than statute. Rules aren’t tightening everywhere at the same pace, but the bar for evidence is rising everywhere. Being able to show continuous oversight is becoming as important as the inspection programme itself.

New technologies and their impact on RBI

Against that backdrop, continuous acoustic monitoring is gaining ground. Technologies like acoustoSENSE pick up the ultrasonic signatures equipment generates in operation, which lets operators catch subtle changes in behaviour as they happen. In practice that’s a retrofit sensor with no penetration of the pressure boundary, certified for hazardous areas, reporting over LoRaWAN into systems the operator already runs. This matters because a monitoring programme that needs new infrastructure rarely survives contact with a capital budget.

For PRDs, that means earlier warning of leakage, lifting events and other abnormal activity that might otherwise go unnoticed until the next inspection. It also means an actual release record i.e. what lifted, when, and for how long. One offshore operator running acoustoSENSE on this basis is targeting a 65% reduction in PSV maintenance costs.

For slow-rotating bearings, it means continuous visibility of developing faults and lubrication issues that conventional monitoring struggles to catch. Catching them early lets maintenance teams plan interventions before they hit production, and it stops unnecessary preventative work being done on assets that are actually fine.

The effect on RBI is substantial. Probability of failure no longer rests on historical trends and inspection records alone, it can be checked against live evidence. Where an asset demonstrates stable behaviour over time, that evidence can support a case for extending inspection intervals. Relief devices need a caveat here, though: published statistical analysis of PRV proof-test data shows that most failures found at test were already present at reinstallation rather than accumulated in service, and jurisdictions vary sharply in how much latitude they allow. Some accept risk-based intervals, others prescribe fixed ones by rule. Continuous data strengthens the case for changing intervals where that’s permitted. Therefore, RBI stops being a static planning exercise and becomes part of a live, responsive integrity system.

That has a direct effect on maintenance economics. Less reliance on routine manual inspection frees up engineering time and cuts disruption to operations. Earlier detection and intervention lowers the odds of unplanned failure. Maintenance work gets more targeted and more predictable, and it tracks actual condition rather than assumed risk.

For European operators specifically, this matters commercially as well as technically. Cefic’s closures and investments radar, published in January 2026, found announced European chemical plant closures up sixfold since 2022 – around 37 Mt of capacity, roughly 9% of European production – with a lack of energy cost competitiveness cited as the dominant reason in 49% of cases. In that climate, integrity spend has to earn its place on cost grounds as much as on assurance grounds, and increasingly those two arguments point the same way. RBI is still the right foundation for managing asset integrity. It just works a lot better when it’s fed by better and more frequent data.

So the question isn’t whether to adopt risk-based inspection, it’s how to evolve it. At Automa 2026, that question sits squarely inside the Digital Refineries, Digital Twins and Autonomous Operations streams. The strategies that work best will combine RBI’s structure with the immediacy of continuous monitoring, turning inspection from something that happens periodically into something that runs all the time.

Continue the conversation at Automa 2026

Whether you’re trying to close the evidence gap on relief device releases, cut maintenance costs, sharpen your RBI programme, or get earlier warning of slow-rotating bearing failures, we’d like to talk about what you’re dealing with and what we’re seeing across the industry.

Throughout Automa 2026, the Tribosonics team will be hosting private meetings in our VIP meeting suite. Book a meeting to talk through how continuous acoustic monitoring can cut maintenance costs, close the evidence gap on relief device releases and feed live asset condition into the systems you already run.