Poor maintenance practices cost: A report issued this January (2025) by the National Audit Office (NAO) highlights a maintenance backlog across key public services, including schools, hospitals, and prisons, estimated at least £49 billion, with the knock-on effects spanning from downtime to serious safety incidents.

Gareth Davies, head of the NAO, said: “Allowing large maintenance backlogs to build up at the buildings used to deliver essential public services is a false economy. Government needs better data on the condition of its operational assets and should use it to plan efficient maintenance programmes to deliver better services and value for money.”

Queue the need for widespread adoption of predictive maintenance, a step above its preventative cousin, that utilises the latest data analytic tools to tailor servicing practices to the specific needs of a building and its operations.

The Evolution of Maintenance; React, Prevent, Predict

Good maintenance practices have evolved significantly over time. At the bad end of the scale is reactive maintenance, fixing equipment only when a problem arises, often after it has failed completely. This approach is often chosen due to its minimal upfront cost, but in most cases, being reactive is a false economy, resulting in unplanned downtime and high repair expenses. In the worst-case scenario, this approach can endanger life.

Preventative maintenance is a definite step in the right direction, aiming to reduce failures through routine inspections and servicing. By sticking to a plan, FMs better manage workloads and costs, reducing the risk of breakdown.

While undeniably more effective than a reactive maintenance, preventative maintenance has its shortcomings; where checks and servicing are based on a theoretical schedule rather than the actual needs of a building and its infrastructure, parts can end up being replaced when they still have useful lifespan. Planned downtime, which is disruptive for building users, is also an integral feature.

Thanks to the advent of data analytics, Internet of Things (IoT) sensors and AI, we are now at a stage where maintenance can be based on real need, using a predictive approach that identifies potential failures before they occur, reducing unnecessary interventions.

 Predictive Maintenance Explained

Predictive maintenance is data-driven, using real-time monitoring and advanced analytics to anticipate equipment failures before they happen. Unlike preventive maintenance, which relies on scheduled servicing, predictive maintenance only occurs when data indicates that a particular asset needs attention, potentially reducing site visits and unnecessary intervention; if everything’s running smoothly, then no action is taken.

‘Over maintenance’ can have negative implications. In closed system HVAC, for example, tampering with the system can lead to oxygen ingress, exacerbating corrosive conditions.

Key components of predictive maintenance:

• Data collection: IoT sensors that continuously monitor asset performance and collect real-time data, such as vibration, humidity and other equipment conditions.
• Data processing: Cloud-based analytics platforms process the collected data, identifying patterns and anomalies.
• Real-time performance tracking: Live dashboards to provide facilities managers with instant insights into asset performance.
• Model development: Machine learning models that predict potential failures based on historical and real-time data.
• Fault notification: Automatic alerts which inform maintenance teams of potential issues before they escalate. No alert means no need for a site visit.
• Model improvement: The system refines its predictions over time, continuously improving accuracy. This is a key area of development thanks to the increasingly wide-spread adoption of AI.

The Current State of UK Maintenance Practices

According to a poll by SFG20 in 2023, at that time, only 20% of businesses were ‘wholeheartedly’ embracing digital-lead maintenance. Two-years on it is likely that figure has risen, but as the NAO’s report demonstrates, despite the clear advantages of predictive maintenance, its adoption is clearly not widespread and some of our most important infrastructure buildings are not yet benefitting from the latest predictive tools.

The reasons behind this are varied, including:

• Cost pressures: Maintenance budgets are often tight, requiring facilities managers to balance efficiency with financial constraints. This is particularly true in the public sector.
• Skills shortages: The industry is facing a growing gap in skilled maintenance personnel, making it difficult to implement and sustain new maintenance strategies.
• Customer satisfaction issues: Equipment failures and maintenance delays impact the tenant experience, reducing overall satisfaction.
• Complex process interfaces: Managing maintenance across multiple systems and suppliers can create inefficiencies and communication challenges.
• Legacy systems: A big issue for our sector is the existing tools currently in-situ that can never be integrated with new sensor technology and/or AI. The investment of ‘ripping it out and starting again’, can be extremely off-putting.
• Resistance to change: Employees may be hesitant to adopt new technologies or alter established workflows.
• Disparate data: Any predictive maintenance tool is only as good as the data it is given. Often, this is spread across multiple platforms and businesses.

 The devils’ in the Data

As highlighted, predictive maintenance software is only as good as the data it has access to; its quality, relevance and quantity is key for accurate analysis of root causes and forecasting failures ahead of time. Increasing data input is a major challenge to any predictive maintenance programme. Sensors must be added to any machines falling under predictive maintenances’ remit, supported by IT infrastructure processes and trained personnel.

As Deloitte says in its paper on Predictive Maintenance, “the more information is available on events to be predicted the better predictions become.”

 Quantifiable Benefits

Implementing predictive maintenance means considerable investment in money and time; if done well, however, the return on this investment should mitigate initial outlay.

According to a Deloitte’s report, on average, predictive maintenance:

• Increases productivity by 25%
• Reduces breakdowns by 70%
• Lowers maintenance costs by 25%
• Increases equipment uptime by 10 to 20%
• Reduces overall maintenance costs by 5 to 10%
• Reduces maintenance planning time by 20 to 50%

Crucially, for buildings on the path to Net Zero, predictive maintenance contributes to sustainability targets:

• Reduced energy consumption, thanks to optimised equipment operating
• Reduced carbon footprints, cutting site visits and associated travel
• Prolonged asset life reduces overall supply chain energy consumption and emissions
• Resource allocation is more efficient. Our BiO® service management platform can be used to further enhance leaner practices by organising and standardising internal and external maintenance teams, ensuring the right engineers with the right qualifications are where they need to be at any given time, preventing wasted site visits.

Successful Predictive Maintenance Implementation

Before investing in any new digital tool, it’s important that the following is considered:

• Define what you want it to do and check any technology functions against this list, make sure you consult with all stakeholders and anyone in your business that’s likely to use it.
• Assess existing infrastructure and identify gaps and areas for improvement.
• Focus on the essentials first and start with small scale implementation, a good way to test key features and prevent initial over-spend.
• Look for software that can grow with the needs of your business and fast-paced digital innovation.
• Keep staff informed and make sure they are trained and supported. Resistance to change is normal but can be mitigated by clear communication and knowledge sharing.

Collaboration is key

Achieving a transformative shift in maintenance requires a collaborative effort from all parties involved: building owner, facilities consultants/buyers, facilities managers and maintenance teams, with a focus on long-term gains, rather than reactive shortermism.

In the case of the public sector buildings highlighted in the NAOs report, maintenance failings not only cost building owners, they cost us too – the £49 billion deficit equates to approximately 4% of the government’s total expenditure in 2023-24, or around £710 for each person living in the UK (based on mid-2023 population estimates).

It’s clear that the days of treating maintenance as a last course of action must come to an end. Well maintained buildings are cheaper to run and less likely to need expensive major repairs therefore reducing the risk of downtime. Well maintained equipment will last longer too; when you add all these benefits together, the cost of doing things the right way is money well spent.

Predictive Maintenance in Action

Leigh Academies Trust (LAT) is one of the UK’s largest Multi Academy Trusts, covering Kent, Medway and South East London, overseeing 32 academies and more than 20,000 students between the ages of 2 months and 19 years-old. Its vast portfolio of buildings, ranging from heritage properties to state-of-the-art new builds, needed a streamlined maintenance approach with quality and efficiency at its core.

Using BIO®, DMA Group has bought LAT’s maintenance operations under one banner, including all building services and other FM functions, consolidating and standardising service delivery across the trust.

What BIO® does for LAT:

• Oversees and keeps track of 6,400 LAT assets across 32 locations
• Manages the teams that maintain them – field-based engineers that cover fabric, air conditioning, gas, plumbing and electrics.
• Manages additional service providers outside of the building services space, including water treatment, lift, drainage, fire and security engineers to insurance inspectors.
• Brings together various maintenance contracts under one single arrangement.
• Ensures standards are kept high as the trust expands, maintaining consistency and value for money.

In Conclusion

The days of treating maintenance as an afterthought must come to an end. A well-maintained building is cheaper to run, less likely to need expensive major repairs, and more resilient against downtime. Equipment that is properly looked after lasts longer, reducing waste and unnecessary costs. When you add all these benefits together, it’s clear that the investment in predictive maintenance isn’t just a luxury—it’s a necessity.

For the public sector, where maintenance failures impact not just budgets but also people’s lives, embracing smarter, data-driven solutions like predictive maintenance is long overdue. With the right technology, the right expertise, and a commitment to long-term efficiency, we can turn maintenance from a cost burden into a strategic advantage.

The tools to make buildings work better already exist—it’s time to put them to use.