Sanne Klit: You Cannot Stock Everything, but You Are Expected to Know What Breaks Next (Part 2)

(Part 2 of our series with Sanne Egholm Klit, Head of Parts Solutions & Warranty at BEUMER)

When operators talk about spare parts planning, the conversation usually starts with a familiar question:

"What do we need to keep on the shelf to avoid downtime?"

For a machine builder, the question is much more complicated.

They are not planning inventory for one site or one production line. They are responsible for thousands of systems, spread across countries, industries and generations of equipment, each with its own usage patterns, maintenance routines and operational risks.

Yet the expectation from customers remains simple: When a part fails, the replacement must already be available. Somewhere.

The installed-base puzzle

Planning spare parts for a single operation is already difficult. Demand is irregular, failures are unpredictable and capital tied up in inventory can quickly escalate.

But machine builders face a different dimension of complexity. Instead of planning for one plant, they must anticipate demand across an entire installed base.

Sanne describes the challenge in very practical terms.

“We look at historical data first. That’s the structured part of the process. But then we also add the knowledge we have in our heads.”

Her team currently performs an ABC analysis every six months, using historical consumption data to identify which parts are most important to keep available. But data alone rarely tells the full story.

A spike in historical demand might reflect a one-time project order rather than a recurring need. A supplier may promise reliable delivery times, yet consistently miss them. Or a customer might mention a possible expansion project months before any formal order is placed.

Those signals matter, as Sanne explains:

“The part about the numbers is structured. But the rest is based on knowledge, what we hear from customers and suppliers, and the information we get.”

In other words: the spreadsheet is only half the model.

The limits of history

This hybrid approach of combining data and experience is common in spare parts management.

Historical demand provides a useful baseline, but it often fails to capture rare failures, new system configurations or sudden shifts in supplier reliability.

And in spare parts planning, those exceptions are often exactly where the highest risk sits.

For machine builders responsible for supporting systems over decades, these low-frequency parts can have an outsized impact. They may move rarely, but when they are needed, they are often business-critical.

You cannot ignore them. But you also cannot stock everything.

Planning under constraints

Even if forecasting were perfect, machine builders still face hard physical limits.

Warehousing space is finite. Capital tied up in inventory must be controlled. And parts stored for years without demand can quickly turn into dead stock. As Reliability Expert Sanjib Das explained in our previous expert blog series, inventory isn’t free: it decays, demands care, and quietly degrades reliability. 

At the same time, the service promise to customers remains open-ended. Some BEUMER systems supported by Sanne’s team have been running for more than 35 years. Sanne says: 

“We have systems that have been running since 1988 and we still support spare parts for them.”

This creates a balancing act unique to the OEM perspective.

Operators optimise spare parts strategy within their own operational context: uptime requirements, maintenance strategy and available budget.

Machine builders must think differently. They must balance demand across multiple customers, multiple generations of equipment, and multiple suppliers, all while maintaining the credibility of their service promise.

The knowledge advantage of machine builders

Despite these challenges, machine builders also hold a unique advantage. They see the installed base. Across airports, logistics centres or mining operations, they observe patterns that individual operators cannot see from a single site.

Maintenance data, component replacements and failure intervals across dozens of systems can reveal early signals about emerging risks. Sanne’s team is beginning to explore exactly this potential, as she explains:

“We are starting to gather more data from sites. For example, when maintenance teams record when they exchange parts.”

That information allows engineers to calculate metrics such as mean time between failures (MTBF) across multiple systems, an insight that can dramatically improve spare parts planning.

One specialist has recently joined Sanne’s team with a specific mission: analysing installed-base data to better predict demand. It is an early step toward something much larger.

From gut feeling to predictive planning

The long-term ambition is clear: replace reactive planning with predictive insight.

If machine builders could combine installed-base data, maintenance records and component lifecycles across their systems, they could begin to anticipate failures before customers even notice them. In other words, the spare part would arrive before the breakdown.

That vision is still emerging.

Today, many decisions still rely on the judgement of experienced planners who understand the quirks of specific systems and suppliers.

But the direction of travel is unmistakable.

The future of spare parts planning will depend less on intuition , and far more on shared data across the ecosystem.

The holes in the cheese

Even with better planning tools, however, uncertainty never disappears entirely.

Suppliers change components. Sub-suppliers discontinue products. Documentation becomes fragmented over decades of projects and upgrades.

Some risks only surface when something breaks.

Sanne uses a simple metaphor to describe this reality. There will always be “holes in the cheese.”

The question is not whether gaps exist, but how quickly organisations can detect and respond to them.

In the next article, we will explore one of the biggest challenges hiding behind those holes: obsolescence.

Because for machine builders supporting systems for 30 years or more, the biggest threat to reliability is not always failure.

Sometimes the part simply no longer exists.
‍

Want to read the whole interview series? 

• Part 1: Spare Parts Is Not a Side Business for Machine Builders
• Part 3: The Obsolescence Trap: Why “Proactive” Is Harder Than It Sounds
• Part 4: The Data Future: Deliver the Spare Part Before the Failure