A tale as old as time: an emergency call in the middle of the night, the panic in the homeowner’s voice, the rush to get to the scene, and the chaos when you get there. It’s a stressful situation for everyone involved, but it’s just the way that things have been done for years. Everything works smoothly until it doesn’t, and then you pick up the pieces.

Preventive maintenance, where it existed, often meant calendar-based inspections and checklist routines designed to reduce the odds of failure, but not eliminate uncertainty. The model has been fundamentally reactive, built around breakdowns, emergency response and time-based servicing.

That model is no longer sustainable.

Energy costs are consuming a growing share of operating budgets, especially in commercial and institutional facilities where HVAC and pumping systems run continuously. Even marginal efficiency degradation — such as a drifting setpoint, a slight hydraulic imbalance or a pump operating off its best efficiency point — compounds into measurable cost. Yet, those inefficiencies rarely trigger alarms under traditional maintenance schedules; they remain invisible until they manifest as complaints, rising utility bills or equipment failure.

At the same time, the skilled labor shortage is forcing contractors and facility teams to rethink how they deploy time and expertise. Reactive service is labor-intensive by definition. It requires troubleshooting under pressure, emergency mobilization and often extended diagnostic time onsite. When experienced technicians are scarce, constantly responding to failures isn't operationally sustainable.

What’s changing is not simply the availability of sensors; it’s the reframing of plumbing and mechanical systems as measurable, managed assets rather than background utilities.

Today issues like pressure transients, flow signatures, temperature gradients, vibration patterns and micro-leaks can be detected in real time. Pumps and valves can be compared to modeled efficiency curves, rather than rule-of-thumb expectations. Even legacy infrastructure can be retrofitted with non-invasive monitoring that provides visibility where none previously existed.

The implications are structural.

Instead of relying on fixed preventive maintenance intervals, service providers can prioritize condition-based interventions. Instead of proving value through completed work orders, they can demonstrate measurable outcomes: verified energy savings, documented uptime, reduced water-loss incidents and extended asset life.

But, have no fear, reactive service certainly won’t disappear — equipment will still fail, seals will still wear, components will still reach end of life. Failure is no longer the primary trigger for engagement, however. We’re seeing a major shift towards prediction, verification and managed performance.

For contractors, manufacturers and facility owners alike, that shift marks more than a technological upgrade. It signals the end of break/fix as the dominant service philosophy, and the emergence of plumbing and mechanical systems as continuously monitored, data-informed infrastructure.

The question is no longer whether predictive service is possible. The question is how quickly the industry adapts to a model where reacting late is the most expensive option available.

From fixed PM schedules to condition-based maintenance

For many years, the most proactive way to prevent breakdowns was to have your systems checked and maintained regularly. However, traditional preventive maintenance (PM) programs, built on fixed intervals, often miss subtle performance drift. Equipment may pass inspection while operating inefficiently for months, and the impact shows up in energy bills long before it appears in a failure report.

Offering Manager, Platform & Ecosystem, Armstrong Fluid Technology, Carlos Chamorro describes the economic driver behind this shift plainly: “Small efficiency degradation on equipment has compounded negative impacts on energy usage, and this typically goes undetected during PM cycles.”

Predictive analytics introduces a different framework. Rather than asking whether a scheduled task has been completed, operators compare live performance against expected performance. Chamorro notes that leveraging digital models “to compare against their actual operation can help building owners and operators understand abnormal vibration, temperature, efficiency, patterns before these become costly reactive repairs.”

This is where model-based validation becomes powerful. Instead of relying on intuition — “it sounds off” or “it feels hot” — operators can benchmark actual performance against high-efficiency sequences of operation and known performance curves.

Chamorro emphasizes that this approach “fundamentally shifts from fixed PM schedules, and reduce[s] unnecessary service to equipment that works properly.”

In other words, analytics prevents both under-maintenance and over-maintenance. Resources are deployed based on deviation severity, not calendar dates.

Vice President, Intelligent Water Solutions, David Benaiges of Watts reinforces the enabling technology: “We can now measure performance in real time with sensors, connected equipment, AI tools and analytics, and that completely changes the equation.” Verification also becomes quantifiable. Chamorro notes, “Operators can measure and verify before and after, providing data-driven evidence that the maintenance addressed the customer problem.”

This “measure and verify” capability aligns closely with ASHRAE Guideline 36 and broader industry moves toward performance validation. Instead of assuming maintenance restored efficiency, operators can document that it did; often saving headaches at a later date.

The most immediate impact of predictive analytics is visible in rotating equipment and central plant systems. Chamorro points to “Pump, fan and motor reliability, particularly identifying bearing wear, misalignment and flow-related stresses.”

Historically, these conditions were diagnosed after vibration became obvious or failure was imminent. Today, continuous monitoring can detect early-stage anomalies in vibration patterns and efficiency drift long before catastrophic failure.

At the plant level, digital twins are expanding predictive capability even further. Chamorro highlights “HVAC plants with digital twin models for chilled-water and condenser-water systems where degradation patterns involve multi equipment and multiple coupling connections.”

In these environments, performance issues rarely originate from just a single component. Chamorro explains, “Rule-based asset management and monitoring support early detection of system-level issues that impact building operations (e.g., low Delta T, hydraulic imbalance).”

Analytics can replace the historical practice of adjusting one component at a time and hoping for improvement. Instead, operators evaluate system-level KPIs.

Chamorro further distinguishes between equipment metrics and operational metrics: “Distinguishing between operating KPIs vs equipment KPIs, measuring the right data can help building owners proactively focus on the right metrics such as energy intensity (KWH/sqft), uptime percentage, tenant retention impact, environmental social and governance (ESG) reporting and carbon intensity, among others.”

This reframing matters. A pump might technically function within acceptable parameters, yet the building could still be underperforming from an energy intensity standpoint. Analytics shifts attention from whether equipment runs to how effectively the system delivers outcomes.

Benaiges echoes this performance-based shift from the plumbing side. Predictive modeling requires establishing correlations: “To move beyond intuition, we need to establish correlations between what we see in the data and actual outcomes. That means looking at multiple inputs, such as indoor and outdoor temperature, humidity, occupancy levels, and even benchmarking similar properties of similar size and use.”

Context transforms raw data into actionable intelligence. A spike in flow or load may be normal under certain occupancy conditions, but without contextual modeling, operators risk misinterpreting normal variance as malfunction.

The foundation of predictability

If reactive service is the legacy model, the foundation of predictive service is systems thinking, paired with measurable visibility.

I spoke with Stephanie Radel of GF about the shift in structural terms, as GF approaches plumbing as a system, rather than a collection of components. “A systems-level approach creates predictability. When piping materials, connection technologies, and distribution architecture are engineered as a unified platform, the system behaves as a measurable asset rather than a collection of independent parts.”

When systems are engineered as complete unified platforms rather than assembled piece-by-piece, variability narrows. Consistent materials, standardized joining technologies and intentional distribution geometry create repeatable performance behavior. As Radel explains, “That consistency allows owners and contractors to establish performance baselines and detect early deviations that signal stress, inefficiency, or emerging failure modes.”

Similar to the importance of quantification? Baselines. They are the critical first step in moving beyond break/fix. Without them, every anomaly looks like an emergency — or worse, goes unnoticed.

The systems approach also changes decision-making during design and specification. “System design anchored in lifecycle performance shifts decision-making upstream. Instead of optimizing for lowest installed cost, stakeholders evaluate durability, performance and serviceability over the time of operation.”

This upstream shift is echoed in the mechanical space. Chamorro points to the impact of several issues compounding together: “Small efficiency degradation on equipment has compounded negative impacts on energy usage, and this typically goes undetected during PM cycles.”

When equipment performance is not benchmarked continuously, degradation accumulates quietly — increasing energy spend long before a failure even occurs. In facilities where energy accounts for a significant share of operating budgets, that drift is no longer acceptable.

Benaiges describes why technology is accelerating the shift: “Technology now makes it possible for us to understand when something needs attention instead of waiting until something needs repairs, or relying on time-based rules. We can now measure performance in real time with sensors, connected equipment, AI tools and analytics, and that completely changes the equation.”

Taken together, these perspectives reinforce a common theme: lifecycle performance must be engineered into the system, not layered on afterward.

Radel connects design directly to maintenance strategy: “This enables predictive maintenance strategies built around known material behavior, validated connection integrity, and integrated monitoring points. The result is a transition from reactive repair to managed asset performance.”

However, monitoring cannot simply be added to a poorly organized system. Architecture determines insight.

“Designing for continuous monitoring requires intentional system architecture. Distribution networks should be zoned to allow meaningful data segmentation, enabling precise fault isolation and performance tracking,” Radel says.

Zoning transforms data into actionable intelligence. Rather than investigating an entire building when an anomaly appears, operators can isolate the issue to a defined branch or zone. Strategic sensing further sharpens resolution: “Strategic placement of sensing nodes, aligned with branches and critical risk points, improves diagnostic resolution and accelerates response time.”

Watts sees similar opportunities in retrofit scenarios. Historically, plumbing systems were either analog or limited to rigid building management systems. Today, Benaiges notes, “virtually any building can be retrofit with lower-cost, non-invasive sensors that are easier to install and deploy in hours instead of days or weeks.”

That flexibility expands systems thinking beyond new construction. Whether through permanent dashboards or temporary deployments, monitoring becomes a practical field tool. As Benaiges explains, remote platforms “become a tool that changes how contractors work and improves productivity. It helps them respond faster, reassure customers when an issue can safely wait instead of requiring an emergency visit, and make better-informed decisions.”

Chamorro similarly highlights how system-level monitoring surfaces issues that would otherwise remain hidden: “Rule-based asset management and monitoring support early detection of system-level issues that impact building operations (e.g., low Delta T, hydraulic imbalance).”

Predictability is the antidote to reactive service.

Closing the gap between data and action

While predictive analytics offers unprecedented visibility, it introduces a new challenge: noise.

False alarms can undermine trust quickly. John Lee, Group Product Manager, Connected Water, of Moen underscores the importance of algorithm quality: “False alarms dilute confidence, so accuracy is critical.”

Connected water systems must distinguish between legitimate anomalies and benign variation. Lee notes that platforms, like Moen’s Flo Smart Water Monitor and Shutoff, “leverage[s] years of nation-wide water usage data to continuously refine the algorithm. The sheer breadth and depth of the data and the quality of the data science behind the product is the difference between an annoying false alarm and a timely ‘money-saving’ alert.”

That principle extends beyond leak detection. In commercial mechanical systems, excessive notifications can overwhelm service teams, leading to alert fatigue. Benaiges highlights the importance of contextual interpretation: “Instead of reacting to every spike, contractors and facility managers can determine whether a pattern is expected or truly problematic.”

Bridging the gap between data and action also requires system mapping. “To turn data into insight, you need to know how the pieces connect and what the implications are when something changes.” Without that understanding, analytics may identify a symptom without clarifying the root cause.

Ultimately, predictive analytics only replaces intuition when it supports prioritization. The goal is not more data; it is earlier, clearer intervention. Chamorro notes, “Predictive insights allow service providers to guarantee uptime, because risks are quantified and monitored and continuously addressed.”

So, quantification is the differentiator. When risk is measured, interventions can be ranked by severity and economic impact. Technicians arrive informed, and service becomes strategic rather than reactive.

Despite these major changes, the importance of intuition does not disappear — but it is increasingly validated, refined and sometimes challenged by data. The result is a service model grounded not in guesswork, but in measurable performance trajectories.

New systems emerging

As predictive analytics matures, the shift from reactive service to managed performance is still finding its rhythm. There are many new options available, and entirely new service models are now becoming viable.

A. Performance-based agreements

Chamorro makes the connection explicit: “Predictive insights allow service providers to guarantee uptime because risks are quantified and monitored and continuously addressed.”

Guarantees are difficult to offer in a reactive model. Breakdowns are unpredictable, diagnostics are time-consuming and outcomes vary. But, when continuous monitoring establishes baselines and identifies early degradation, risk becomes measurable rather than speculative.

That measurability underpins outcome-based contracting. Chamorro notes that “Performance-based contracts become feasible when both parties share the same set of outcomes with transparent ESG and reliability targets.”

Instead of billing for labor hours or completed tasks, service providers can align compensation with defined KPIs — energy intensity, uptime percentage, Delta T performance or verified savings. Then, both parties operate from the same dataset, and maintenance success is validated rather than assumed.

He adds that, “Shared-risk and share-the-wallet agreements work because decisions are made using measurable and verifiable data, to validate the outcomes, not checklist for tasks completion.”

The emphasis shifts from activity to results. The service relationship evolves from transactional to performance-aligned.

B. Subscription and hybrid maintenance models

In residential and light commercial markets, always-on water monitoring is opening the door to subscription-based maintenance. Lee acknowledges that new models may take time to scale, but the infrastructure is largely in place: “Most of the enablers are within clear reach for this transition to materialize.”

Connected platforms provide real-time alerts, usage profiling and remote shutoff capabilities. That persistent visibility supports recurring service arrangements, rather than episodic emergency calls.

Lee anticipates experimentation before stabilization: “While I can’t look into the future, I naturally expect a hybrid model to surface as different business models are tested and optimized.”

Hybrid models may combine traditional service retainers with performance monitoring dashboards, leak detection subscriptions, or usage-based advisory services. For contractors, this represents a shift from revenue driven by failure to revenue driven by prevention.

Across contributors, a consistent theme emerges: predictive analytics enables continuity. Service is no longer defined by isolated events, but by ongoing oversight.

Chamorro summarizes the transformation: “The role shifts from ‘fix it when it breaks’ to continuous diagnosis and advisement for building owners and facility managers.”

That shift elevates contractors from responders to reliability partners. It also aligns plumbing and mechanical service with broader asset-management practices already common in industrial and data-center environments.

One thing is certain: reactive service will not disappear. Components will still fail, and aging infrastructure will still require replacement. But, the industry’s center of gravity is moving.

Technology adoption, rising energy costs and labor scarcity are converging to make unpredictability expensive. As Benaiges notes, “With limited labor, contractors simply cannot afford to be constantly reacting.”

Systems engineered as unified platforms, monitored continuously and benchmarked against digital models are redefining what maintenance looks like.

The commercial implications follow naturally. When outcomes can be measured, contracts can be structured around them, and service becomes strategic rather than urgent.

The end of reactive service is not about eliminating breakdowns; it’s about eliminating surprise as the primary trigger for engagement.

In a data-informed environment, plumbing and mechanical systems become managed infrastructure — visible, measurable and continuously optimized. Contractors who embrace that reality position themselves not simply as technicians, but as long-term asset stewards in an increasingly performance-driven built environment.

When plumbing and mechanical systems are engineered as measurable platforms — digitally visible, performance-benchmarked and accessible by design — they move from hidden utilities to managed assets. And, once they are managed assets, reactive service becomes the exception, rather than the operating model.