The sounds of equipment carry meaning: How reliability leaders turn data into action
Key Highlights
- The key to advancing predictive maintenance is not more data, but better listening and interpretation of signals from equipment.
- Standardized work processes, clear communication, and leadership support are critical for scaling and sustaining reliability programs.
- Building a strong reliability culture ensures that technology investments translate into meaningful operational improvements.
- Effective internal communication of reliability successes helps justify investments and fosters organizational support.
Industrial equipment has always been talking, and we’ve been listening and collecting. The challenge is taking all those signals and translating them into better maintenance decisions. For a long time, predictive maintenance has focused on finding equipment problems earlier. As condition monitoring technologies mature, so can operational decision-making, if manufacturer’s practices mature along with the technology.
Opening Ultrasound World 2026, Blair Fraser, executive vice president, and Scott Stephens, director of GTM strategy and enablement, at UE Systems, argued that the next evolution in predictive maintenance is not about collecting more or better data. “The data is accessible to us. We have matured in that category,” Stephens said. “The real opportunity now is to make sense of all this data.”
If you can listen closely enough, your equipment is telling you when to act before failures occur. “We do not need to hear more. The world is full of noise. We need to listen better,” Fraser said. This theme rang loudly throughout the conference—the difference between hearing and listening is the difference between empty data and meaningful action.
Throughout the keynote, the two UE executives portrayed a PdM future built around three connected steps: detect, diagnose, and act. Artificial intelligence will also play a role here to reduce a big predictive maintenance bottleneck around interpreting the growing volumes of condition monitoring data. “This is about prioritization,” Fraser said. The industry’s ability to collect that data advanced much more rapidly that our ability to work with that data.
Automation can also close the reliability loop by triggering, for example, automated lubrication and directing technicians to assets that need attention first. AI and automation are intended to amplify reliability professionals’ expertise by finding the signals in the noise that point toward decisions and actions.
A pre-recorded video narrator outlined the vision: “We believe industry should listen differently because the sounds of equipment carry meaning. Before waste, before downtime, before failure, there is a signal. That signal can reveal loss, friction, leakage, abnormal behavior. And when those signals are understood, they become intelligence. Intelligence that can drive action.”
Across customer case studies, technical sessions, and panel discussions, speakers returned to the same themes: predictive maintenance succeeds when organizations build consistent processes, respond to condition monitoring insights, and establish strong reliability fundamentals before adopting AI. This creates cultures that sustain reliability practices and clearly communicate the business value of maintenance.
What surprised me most about the conference content was how much of the discussion was not centered on ultrasound hardware or software. Instead, speakers repeatedly returned to management disciplines—standardization, communication, accountability, and training. Many different industries described very similar obstacles.
1. Reliability programs fail because of inconsistency, not technology.
The biggest barrier to scaling predictive maintenance is not access to sensors, software, or analytics. It is inconsistent execution across technicians, sites, and departments. Many organizations have invested in condition monitoring but struggle to create repeatable processes. Standard work, clear expectations, and consistent data collection determine whether reliability programs scale.
“It’s not a capability problem; it’s a consistency problem.”
— Jeremy Bey, strategic accounts leader at UE Systems.
2. Data only creates value when maintenance teams act on it.
Condition monitoring generates information, but reliability improvements happen only when those findings become action in the form of work orders, repairs, engineering changes, and operational decisions. A predictive maintenance program without a response process simply creates a larger backlog of unresolved issues.
“You can have all this great technology...but if you don't have people on-site that care enough to go out and fix something or go look at it and diagnose it, then you still end up with a failure.”
— Brian Heinsius, senior reliability manager at Eco Material Technologies.
3. AI will amplify reliability — but cannot replace fundamentals.
AI can identify patterns, prioritize risks, and combine multiple condition monitoring inputs, but only if organizations have reliable data and mature processes. Plants investing in AI-driven maintenance need to first establish strong data governance, asset histories, inspection standards, and failure analysis practices.
“AI accelerates a strong reliability process, but it is not a replacement.”
— Chris Hileman, reliability engineering manager at Amazon.
4. The reliability culture determines whether technology survives.
Deploying predictive maintenance tools is easier than sustaining their use. Reliability must become part of daily work, not a special initiative. Workforce turnover and changing priorities can quickly undermine reliability investments unless practices become embedded into the organization.
"Build it, make the process as if you're not going to be there tomorrow. That way, it's sustainable and it continues on even after we're gone."
— Nikolas Allen, junior reliability engineer at Irving Tissue.
5. Maintenance teams need to sell reliability internally.
Reliability improvements are often invisible because the failure never happens. Teams must communicate avoided downtime, cost savings, safety improvements, and business impact. The ability to demonstrate value influences future capital investment, leadership support, and organizational perception of maintenance.
“That’s not what maintenance does. We don't sell it. I found out one of the key contributing successes was communication, and so we needed to communicate these findings.”
— Eric Holt, technical training instructor at Michelin.
The next generation of predictive mainteance
The biggest takeaway for me from Ultrasound World 2026 wasn’t about predictive maintenance entering the AI era. No one is questioning whether AI fits nicely with PdM. Of course, it does. The question is whether manufacturers have built reliable enough infrastructure and processes to benefit from that partnership. The industry’s definition of mature, as far as conditioning monitoring, is changing. Mature programs aren’t identified by sensor deployments and analytics sophistication. The consistency with which organizations respond to the information, how they understand their data, and how they prioritize and act on insights matters much more.
The technology continues to evolve rapidly. But the speakers who shared successful implementations kept emphasizing the same fundamentals: standardized work processes, clear communication, leadership support, and a culture that turns condition monitoring into action. Luckily, those aren’t new ideas for reliability, and the same foundation can deliver on the next generation of predictive maintenance.
About the Author
Anna Townshend
head of content
Anna Townshend is the chief editor and head of content for Plant Services. She has been a journalist and editor for more than 20 years and joined Plant Services and Control Design as managing editor in June 2020. Previously, for more than 10 years, she was the chief editor of Marina Dock Age and International Dredging Review. In addition to writing and editing thousands of articles in her career, she has been an active speaker on industry panels and presentations, as well as host for Great Question: A Manufacturing Podcast. Email her at [email protected].