Author Copperberg Editorial Team | *This article was developed using a combination of human expertise and AI-assisted writing. The concept, structure, and editorial direction were defined by our team, while elements of the text were generated with the support of advanced language tools. All content has been reviewed, refined, and approved by humans to ensure accuracy, clarity, and relevance.

Across capital-intensive industries, the economics of rotating equipment are shifting. The traditional focus on selling robust machines, spare parts, and scheduled services is no longer sufficient in markets where customers now benchmark suppliers not only on reliability, but on total process performance, operating cost, and environmental impact.

At the Field Service Forum UK 2026 – Power of 50, Burak Özkök of Alfa Laval explored how this shift is playing out in practice, moving beyond asset health monitoring towards optimisation of the processes those assets support, and evolving from transactional maintenance into performance-based service models that create both greater risk and greater upside for OEMs.

From catastrophic failures to continuous insight

High-speed rotating equipment, such as centrifugal decanters, operates under extreme mechanical stress, with rotational speeds generating forces of several thousand g. Across industries, including wastewater treatment, mining, and bioethanol production, failures are infrequent but highly consequential. When breakdowns occur, they can trigger unplanned downtime, major damage to critical components, and repair costs that approach a significant proportion of the machine’s replacement value.

These risks have been managed through preventive maintenance, such as fixed overhaul intervals, scheduled inspections, and predefined service calendars. However, operational experience has exposed the limitations of this approach. Even equipment with a long history of reliable performance can suffer unexpected failures between planned service events, leading many operators to question why conventional maintenance regimes fail to detect emerging problems earlier.

This has accelerated the shift towards continuous condition monitoring. Instead of relying solely on periodic inspection, networks of sensors continuously track the condition of bearings, motors, belts, and other rotating components in real time. Data is analysed locally and in the cloud, allowing algorithms to identify early indications of deterioration through subtle changes in vibration patterns and related operating parameters.

The objective is to move from static, calendar-based maintenance towards dynamic, condition-based decision-making. Maintenance interventions can be delayed when equipment remains healthy or accelerated when warning signs emerge. For operators, this offers the prospect of lower unplanned downtime and improved asset utilisation. For OEMs, it creates the foundation for more advanced and performance-oriented service models.

Risk, revenue, and the logic of maintenance

Once continuous insight into machine health is established, service models can move beyond one-off inspections and spare part sales, facilitating long-term agreements where the OEM assumes responsibility for keeping the equipment available, often for a fixed annual fee.

The logic is attractive to customers, as paying a predictable amount each year transfers the complexity of maintenance planning, condition analysis, and intervention timing to the supplier. The OEM, equipped with condition data and domain expertise, optimises interventions to balance operational risk and cost while protecting the asset.

However, this model is not trivial for the manufacturer. It reshapes the risk profile. If a major failure still occurs, the customer continues to pay the fixed fee, but the OEM absorbs the repair cost, potentially wiping out margins for that contract period. The reward is the prospect of recurring, higher-value revenue streams, but it demands confidence in both technology and analytics, as well as a deep understanding of asset behaviour over time.

This is a critical step on the servitization journey, moving from selling parts and labour to selling assurance. It also exposes a limit. Even the most advanced condition monitoring and uptime guarantees remain anchored in a break-fix logic. The focus is still the machine, not the performance of the process that the machine serves.

From asset health to process performance

Condition monitoring is increasingly extending beyond the health of individual assets and into the optimisation of the wider process that those assets support.

In many industrial environments, operating conditions are highly variable. Feed quality, production loads, environmental factors, and upstream processes can all fluctuate significantly over time. Historically, operators have relied on periodic measurements and manual adjustments based on experience.

With greater use of sensors, connectivity, and real-time analytics, processes can now be monitored continuously rather than intermittently. Data from multiple points across a system provides ongoing visibility into both input conditions and output quality, allowing operating parameters to be adjusted dynamically as conditions change.

Once connected to automated controls, this enables systems to respond in real time, adjusting throughput, dosing, energy use, or other process variables to maintain performance and efficiency.

At this stage, the focus moves beyond traditional condition monitoring. The objective is no longer simply predicting when a component might fail, but continuously optimising operational outcomes such as quality, efficiency, throughput, energy consumption, and cost. Assets become part of a responsive, data-driven process rather than isolated machines maintained on fixed schedules.

The economics of optimisation, from thousands to tens of thousands

Once process performance becomes the focus, the economic model changes significantly.

Traditional service and maintenance costs for industrial equipment are often modest compared with the much larger operational costs surrounding the process itself, including energy consumption, consumables, waste handling, logistics, compliance, and efficiency losses.

A process-optimisation approach targets this broader cost base. By using real-time data to stabilise operations and adjust performance dynamically, organisations can reduce unnecessary consumption, improve output quality, minimise waste, and make better use of existing assets.

The most important shift is that value is no longer created primarily through avoiding equipment failure or extending component life. Instead, it comes from improving operational efficiency and process outcomes across the wider system.

This changes how OEMs position themselves. Rather than being viewed mainly as equipment suppliers or maintenance providers, they become partners in operational performance, helping customers improve productivity, sustainability, and cost efficiency over time.

Experts, not algorithms, are the first barrier

While technology and economics support the move to process optimisation, the internal journey inside industrial organisations is rarely smooth.

A recurring pattern is resistance from domain experts. The process is seen as too complex, too variable, and too dependent on tacit operator knowledge to be captured and controlled by algorithms and standardised sensors. Initial attempts to model or control the process often run into unexpected behaviours and edge cases, seemingly confirming sceptical views that it cannot be done reliably.

Once a system demonstrates that it can respond appropriately to changing conditions, the nature of the discussion begins to shift. Resistance based on feasibility tends to give way to more constructive scrutiny of details such as parameter settings, edge cases, and operating assumptions.

At that point, operational expertise becomes a critical input rather than a counterweight to the system. Deep process knowledge starts to shape and refine the model, helping ensure that optimisation reflects real-world conditions rather than theoretical expectations.

For organisations pursuing this direction, the key challenge is often less technical than organisational. Building confidence among experts requires time, transparent results, and iterative refinement based on real feedback. Without that trust and engagement, even promising optimisation initiatives risk remaining limited in scope rather than scaling into robust, widely adopted services.

The customer paradox when becoming digital

Industrial operators want to use data better and become more digital, yet many are simultaneously reluctant to allow data to leave their plants. Concerns about cybersecurity, competitive sensitivity, and loss of control frequently collide with the recognised potential of cloud analytics and connected services.

This paradox has direct implications for how optimisation and monitoring solutions are designed and commercialised. Customers expect sophisticated insights, often benchmarked against what competitors can offer, but may resist the architecture that would enable it at scale.

Two strategies appear particularly relevant in such an environment. First, maintaining flexibility in system design, for instance by offering data exports via APIs into customers’ preferred supervisory systems or platforms rather than forcing adoption of vendor-specific portals. Second, building trust through clear data governance, demonstrable security, and tangible value cases that justify any perceived risk of data sharing.

In many cases, practical value, such as measurable savings in polymer, energy, or disposal costs, demonstrable compliance improvements, or reduced downtime, is the most effective counterweight to data anxiety. When the economic argument is strong enough, reluctance tends to soften, particularly in competitive or highly regulated segments.

From component supplier to performance partner

Where companies once supplied discrete pieces of equipment, the focus is increasingly shifting towards delivering process-level outcomes. The emphasis moves from individual machine performance to sustained quality and operational efficiency, enabled by sensing, analytics, and control.

This shift raises several key considerations:

• Condition monitoring is necessary but not sufficient for servitization. It enables new maintenance models and some recurring revenue, but the real value lies in optimising what flows through the machine.
• Process optimisation requires cross-disciplinary capability. Mechanical engineering must be combined with process expertise, data science, and control knowledge. Partnerships with sensor and software providers are often essential.
• Risk must be managed, not just transferred. Performance-based models shift risk to the supplier, but diagnostics and analytics allow it to be actively understood and controlled.
• Buy-in is critical. Experts must trust the system reflects real process behaviour, and customers must see clear, measurable value.
• Autonomy is the direction of travel. Assets increasingly self-adjust, optimise performance, and detect issues with limited human input. The aim is not to replace operators, but to free them for higher-level decisions.

In markets where customers are no longer satisfied with machines that just run, but expect systems that continuously deliver optimised outcomes, these decisions will increasingly define who leads and who follows.