Maintenance is where real risk resides

We tend to treat maintenance as background noise – and pretend that it is not interesting.

Something operational that’s only discussed when things go wrong. Something that just needs to keep things going while attention moves to more visible and attractive priorities. But that is part of the problem.

What we are really managing is institutional technical debt. Years of underinvestment, short-term fixes, and decisions that made sense at the time but have accumulated into something more structural.

You don’t see it. Until you do. And then it’s an under-pressure estates department’s responsibility to fix.

The estate we have isn’t always the one we need

Most universities are operating buildings that were never designed for how they are now used. Patterns of use have shifted. Space is more variable. Research and power consumption is more intensive. Expectations around comfort, reliability, and digital integration are higher than they were even ten years ago.

But the underlying infrastructure hasn’t kept pace. We are asking more from assets that are already beyond their optimal lifecycle. Then we act surprised when they fail more often, cost more to run, and become harder to maintain.

That gap between expectation and reality is where the pressure sits. It shows up first in maintenance teams, but it doesn’t stay there. The maintenance teams are often seen to be the problem rather than being seen as the diligent few keeping things going.

Deferred maintenance is a series of decisions

Backlog maintenance is often presented as a number. A liability sitting somewhere in a report. In practice, it’s a set of trade-offs. Do we replace that item of infrastructure now, or keep it going for another year? Do we invest in condition data, or accept that decisions are being made with partial information? Do we prioritise compliance, performance, or user experience when we can’t do all three?

Every institution is making these calls. The issue is not that they are being made, it’s that over time they accumulate. And once they accumulate far enough, the cost of recovery is no longer incremental. It becomes structural. They move from maintenance issues and become condition issues.

One of the more interesting differences across functions is how the issue is framed. IT colleagues talk about legacy systems and the need for continuous investment to maintain performance and resilience – while in estates, we tend to talk about maintenance, backlog, and reactive versus planned activity.

But it is the same underlying problem. Ageing infrastructure. Increasing interdependencies. Rising expectations. Constrained resource. It’s just over two different timescales and different costs. Though the issue is compounded when they coincide. When a building system fails, it can affect digital systems. When digital infrastructure is constrained, the estate often does not function as intended. These are not separate challenges. We are often managing one system, even if we mostly organise ourselves as two.

The impact goes well beyond cost

It’s easy to reduce maintenance to a financial discussion. Reactive versus planned spend. Efficiency. Value for money. But the consequences are broader.

It shows up in student experience when teaching spaces do not perform as expected. It affects staff productivity when systems are unreliable and senior leaders are inundated by complaints from our academic colleagues. It directly impacts research where environmental control and resilience are critical. And it sits at the heart of sustainability because inefficient assets are almost always more carbon intensive.

There is also a quieter risk around resilience. The systems that cannot fail are often the ones under the most pressure.

For estates teams, none of the above is abstract. We are often working with incomplete asset data. The best we have at a sector level, the AUDE Estates Management Report, is published annually, and for those teams using it to actively research and benchmark against peer institutions, is an immensely useful data background.

We are managing systems that have evolved over decades rather than being designed as a whole. We are operating within supply chains and contract models that do not always align with how we want to deliver services. And we are doing it with well-documented skills shortages in technical roles (at the moment, it’s a nightmare trying to recruit quality people). At the same time, expectations are increasing. More assurance. More insight. More performance. More data. The shift from reactive to planned maintenance is widely understood. Delivering it consistently, at scale, is the difficult part.

We are still designing future problems into new buildings

There is also an uncomfortable truth in how we approach new development. Projects are often optimised for capital cost and delivery timelines (time, cost, quality PM triangle). Whole-life performance and maintainability are considered, but they are not always the deciding factors. That shows up later. Cheaper specifications that are difficult to access or maintain and more expensive over lifespans. Inconsistent components that increase operational complexity and downtime/disruption. Limited asset information at handover.

We then expect these buildings to perform better than the ones they replaced, without necessarily changing the underlying approach. There is a stronger role here for maintenance teams to challenge design decisions earlier, not after the building is handed over. They just don’t have the resources as they’re firefighting live issues and work demand.

Estates functions are increasingly being asked to operate at a more strategic level. To translate estate condition into institutional risk. To inform investment decisions. To align with wider priorities around sustainability, student experience, and financial resilience. That shift is necessary. But it requires maintenance to be understood differently at leadership level. This is not just about fixing things. It is about sustaining the performance of the institution’s physical infrastructure in the same way IT sustains its digital infrastructure. Both are critical. Both require ongoing investment. Neither can be deferred indefinitely without consequence.

The backlog is growing

The sector needs to be more honest about this. We have been very good at investing in new buildings when the budget was available. They are visible. They signal ambition. They are easier to align with growth narratives. We have been less consistent in how we talk about, report, and prioritise the condition of what we already have. Maintenance is often something to manage within constraints, rather than something that requires deliberate strategic focus. At some point, that becomes a choice.

The question we are avoiding is that every institution is operating within financial pressure. That is not going to change. But the longer-term question is whether we can continue to carry the level of technical debt, or condition backlog, that is building across both estates and IT.

Because it does not stay static. It grows. And when it starts to unwind, it rarely does so at a convenient time or at a manageable cost. So, the question is not whether we can afford to invest more in maintenance. It is whether we can afford not to.