We spoke to FMJ this month about how facilities professionals can invest in AI with greater confidence. See advice our Head of Global Development, Anthony Clark, had to offer.
According to the ONS, there is strong operational optimism regarding AI adoption in the UK FM sector but cautious, limited implementation. Research by MRI suggests cost pressures remain a barrier. How can FM justify the return on investment?
Till Eichenauer, askporter CEO believes most ROI cases fail because they’re built on benefits nobody can measure. Efficiency gains and better visibility won’t survive the next budget round he argues.
“What does survive is a number the finance director can check. Every operation already pays for the gap between someone spotting a fault and a work order being created. It pays for repeat visits when an operative turns up without the right information, for SLA penalties, and for the admin hours sunk into every job. Add those up, and you have the real baseline.”
Paul Scott, CTO at Matrix Booking advises stressing the advantages of having a greater understanding of how workplaces are being used. By using software to provide data on bookings, occupancy and space utilisation, organisations gain a clear picture of workplace demand, helping them reduce wasted space, improve usage and make more confident estate decisions.
Addressing a measurable operational problem such as excessive energy consumption, repeated asset failures, compliance administration, engineer travel or avoidable downtime, will provide a clear baseline against which to measure whether the investment is delivering value, advises Anthony Clark, Head of Global Development at SWG (Service Works Global).
Don’t overlook the human element. Joshua Greibach, CEO, and Tom Wilcock, COO and co-founders of expansive, caution that the biggest hidden cost in AI investment is not financial, but securing team buy-in and the time and money leaders must invest. They argue that AI must offer more than the occasional useful insights: it should genuinely make day-to-day work easier, rather than simply looking impressive.
Dealing with data
Having access to realms of data can be daunting, particularly when faced with data overload. One of the reasons says Clark is that many teams lack a complete asset register and get overwhelmed by alarms and readings from operational systems. As more sensors are installed, teams may be tempted to collect data indiscriminately. But the solution is not more information. Data should be collected only when it supports a defined operational decision, control, compliance requirement or performance outcome.
Another issue is transparency. An askporter survey revealed that 83 per cent of FM teams lack real-time visibility of their workload.
“They don’t lack data. They lack usable data,” says Eichenauer. “The fix sits at the point of capture. When a task comes in through a structured route, say a QR code on the asset or a short set of guided questions, the location, asset, category and priority are attached from the start. Nobody must tidy anything up later, and reporting becomes a by-product of work that has already happened.”
The survey also revealed that 94 per cent of FM teams ran multiple disconnected systems resulting in fragmented data. This problem however is being addressed says Clark.
“Modern platforms like Geminus allow organisations to retain the best systems for each function while connecting their information through open interfaces, common services and shared building context.”
Worth the investment
Teams that hold off on investing in AI might find they are losing the edge against their competitors warn Greibach and Wilcock. AI is already generating work orders and helping diagnose faults, so teams spend less time buried in admin. Automated invoices, quotes and document handling cuts down on mistakes and helps people make faster, better-informed decisions.
A recent, expansive study showed that when vendors used AI to generate unique quotes, only 3.5 per cent required further adjustments; the rest were viable after the initial AI assessment. This means users can review contractor quotations in minutes rather than manually assessing pricing.
According to Scott, one of AI’s biggest strengths is helping organisations understand how their estate is performing. By analysing workplace data, it can identify trends, predict future demand and recommend opportunities to optimise space, reducing operational costs and identifying the best performing spaces to replicate across the estate.
Sustainability is also a key driver, says Clark: “Not only because organisations are under increasing pressure to reduce carbon emissions and meet net zero targets, but also because of the financial savings this will bring through more strategic maintenance decisions and automating operation based on occupancy or conditions.”
AI is also having a huge impact on field communications says Eichenauer.
“When it comes to field communications where details might be missed on the phone, jobs miscategorised or the wrong trade sent, closing the gap between ‘something’s broken’ and ‘a categorised, assigned work order exists’ is where AI-powered triage has already been proven.
“We see it daily. A visitor spots a broken escalator, scans the wall-mounted QR code, describes the fault in a couple of messages, and a categorised work order reaches the right team in under a minute.”
AI risks
FMs need to consider how AI will be introduced and governed. Clark warns that unmanaged use of public AI tools can create information security, governance and accuracy risks, particularly where employees enter organisational or customer information without approved controls.
AI should not replace a human’s professional opinion, but viewed as a decision-support tool, with human expertise remaining essential for making the final call. The level of human oversight should reflect the level of risk. Low-risk administrative processes may be suitable for greater automation, while decisions involving safety, compliance, significant expenditure or changes to critical building systems should require clear approval. FMs should also expect AI recommendations to be traceable to their source information, permission-aware and auditable.
You’ve also got to interrogate the data as AI-powered recommendations are only as good as the asset records underneath them says Eichenauer: “Bad data doesn’t announce itself; it produces plausible outputs that happen to be wrong. Sort the sources before you automate anything that depends on them.
“Ask for an audit trail of automated decisions, where your data is hosted, and how the supplier meets data protection rules and the AI-specific regulation. A simple test: if you couldn’t explain a decision to your client afterwards, it shouldn’t be automated yet.”
Greibach and Wilcock warn that when rolled out badly, AI feels like an added burden rather than a useful tool. If teams see it as another problem to solve, they will revert to familiar methods and stop using it. Done well, however, AI should feel intuitive, with automation freeing up valuable time. It is worth working through the initial learning curve to realise those benefits.
Future possibilities
Our experts believe that the convergence of connected platforms, digital twins and AI rather than any one of those technologies in isolation will revolutionise FM.
“A digital twin provides the structured context: what an asset is, where it is, what it serves and how it relates to the wider building” says Clark. “Operational systems provide its current and historical condition, while AI helps people interrogate that information and translate it into decisions and actions.”
Greibach and Wilcock predict that AI agents will enable companies to retrieve instant answers to help repetitive tasks and troubleshoot issues across departments. They will also help solve ‘pain points’ around asset lifecycles because once an engineer installs an asset, it’s added straight to the register along with its condition data, operating manuals, replacement costs, planned maintenance requirements and end-of-life expectancy. Rather than cutting people out of the loop, these agents simply take care of the busywork so teams can focus elsewhere, helping to streamline workflows.
According to Eichenauer, natural-language systems could have a huge impact on FM. “For 30 years, the sector has asked people to adapt to its software, through portals, logins, forms and training. That’s why so much of the industry still runs on inboxes and spreadsheets.
“Natural-language systems reverse the direction of travel: an occupant, an operative or a client reports, updates or asks in plain words, on a channel they already use, and structured data comes out the other end. The adoption barrier that has dogged FM software for decades dissolves.”
As Scott concludes: “Ultimately, the future of FM isn’t about adopting more technology; it’s about making existing technologies work together more intelligently. Organisations that can turn connected data into actionable workplace intelligence will be the ones that unlock the greatest value.”
Key takeaways
- AI investment in FM should be justified against measurable operational outcomes such as reduced downtime, lower energy use, fewer repeat visits and less administrative effort, rather than vague efficiency claims.
- The biggest challenge is not a lack of data, but poor- quality, fragmented and disconnected information. Connected platforms and structured data capture provide more valuable insights than simply collecting more data.
- AI is already delivering practical benefits by automating work orders, analysing workplace performance, improving field communications and supporting sustainability initiatives.
- Successful AI adoption depends as much on employee buy-in, training and ease of use as it does on the technology itself. Solutions should simplify day-to-day work rather than add complexity.
- Human oversight remains essential. AI should support decision-making rather than replace professional judgement, particularly for safety-critical, compliance and high-risk operational decisions.
- High-quality asset data, transparency and robust governance are fundamental to trustworthy AI. Organisations should expect clear audit trails, secure data handling and explainable recommendations.
- The future of FM lies in combining AI, connected platforms and digital twins to create intelligent, data-driven buildings that improve operational efficiency and workplace performance.
