Artificial intelligence (AI) is reshaping industries across the globe. In the world of facilities management (FM), where operational efficiency, occupant satisfaction, statutory compliance and sustainability intersect, AI promises much, yet as a sector, FM has been fairly slow on the uptake.
When we conducted research in 2021, 77% FM professionals admitted that FM is ‘behind the curve when it comes to adopting smart technology’, with only 27% at the time unlocking the full advantages of smart tech in business process automation. Fast forward to 2025, and our most recent report revealed at the Workplace Futures conference, showed things are changing. 66% of respondents have AI in their 2025 budgets, but many are still hesitant due to expertise gaps and ROI uncertainty. Barriers include a lack of internal expertise, budget constraints and concerns about data security.
Despite misgivings, AI has a lot to offer, with automation of business processes and workflows leading to much greater efficiencies, saving time and money while improving end-to-end visibility using live data. What’s key is that any digital transformation, with or without AI, is managed and implemented in the right way and using the right skills as it isn’t a quick fix for everything.
Too often businesses invest in software without first understanding exactly what problems they want to solve and what their technology needs to do, or how their organisation must prepare.
Here are some of the common pitfalls:
1. Incomplete or inaccurate data
AI is only as smart as the data it learns from. And the reality is that any AI solution needs lots of high quality data if it’s to make a lasting difference.
In FM, the data landscape is fragmented at best. Multiple legacy systems, inconsistent reporting standards, siloed departments and service partners all contribute to a lack of clean, live, structured information. The result? AI is trained on flawed inputs, leading to faulty outputs. For instance, a machine-learning model might identify a pattern in energy consumption and suggest a change in HVAC scheduling. But if the data ignores factors like temporary occupancy surges or outdated sensor readings, the recommendation can do more harm than good.
We must begin with robust data governance. FM leaders need to treat data as a strategic asset, curated, contextualised, and continually validated. Only then can AI begin to add value, drive productivity gain and enable us to act more quickly.
2. Lack of context
One of AI’s greatest limitations in the built environment is its inability to understand why something is happening. Machines are fantastic at pattern recognition, but they struggle with nuance. Without context, AI can’t tell the difference between an anomaly and a real issue.
That’s why AI in FM must remain a tool, not a decision-maker. A hybrid approach, where machine logic and human judgement work together, is the real future of intelligent building and maintenance management.
3. Legacy systems that aren’t fit for the future
Some older Computer Aided Facilities Management (CAFM) and Building Management Systems (BMS) are not compatible with AI, and for businesses that have these systems but want to move forward, investment in ‘starting again’ is probably the only option. Trying to fit a square peg in a round hole will only cost more in the long run.
This can be achieved slowly, however, so rather than chasing full process automation, FM firms can take a phased approach. Prioritise critical systems and processes where AI can deliver the biggest ROI—like better planned and predictive maintenance for equipment, smart energy optimisation or reduced administrative burden (more about that later) —and expand from there. Hopefully, the savings made by these ‘quick wins’ will help fund future investment, whilst also allowing systems to be tried, tested and refined.
4. Forgetting the ‘human touch’
No matter how advanced AI becomes, it can’t replicate the human experience or original thinking. In FM, statutory compliance and customer service are everything. Customers value trust, and accountability; qualities that can’t be automated. Long term customer relationships are forged on more than business acumen.
5. The cost of AI
AI isn’t cheap. Between the cost of sensors, infrastructure upgrades, software licenses, and skilled leaders and staff to manage it all, the investment is significant. But the benefits: greater productivity, greater efficiency, reduced downtime, better energy efficiency, and improved occupant satisfaction, will all reap dividends overtime.
Many of these benefits fall to the end user, which begs the question, who should pay for AI? Should it be the customer, seeking long-term savings and compliance? The service provider, looking to differentiate in a competitive market? Or should the cost be shared, perhaps built into performance-based contracts?
FM firms need to be transparent about the costs and benefits of AI initiatives. Business cases must be tailored, showing clear payback timelines and KPIs.
But FM firms must also recognise that there is much they should be doing anyway to get their own house in better order. Customers can help by structuring commercial contracts with terms and conditions that recognise, value and incentivise the investment their suppliers make into technology, rather than the staid and traditional contracts that haven’t changed in decades.
Our industry typically operates on very low margins, so expecting supply-side to do everything is neither feasible nor sustainable.
6. Ethical concerns
The use of AI brings up important ethical concerns related to data privacy, bias, and accountability. FM companies need to assess how AI may affect employee roles. There’s a risk that it could unintentionally support discriminatory outcomes if the training data is biased. For instance, Amazon discontinued its use of AI in recruitment after the system began automatically rejecting female applicants.
To implement AI ethically, organisations must prioritise transparency, fairness, and ongoing evaluation to ensure the technology functions as intended and avoids harmful side effects.
And finally…
7. Not understanding the problem before you try and solve it
Any investment in digital transformation must begin with understanding the problem/s. Speak to everyone in your business, evaluate what’s working and what isn’t, audit assets and working practices, identify the quick wins. We did this within DMA before developing our own workflow management software, BIO®. By consulting teams across the business we got a feel for their pain points and possible areas for improvement.
Before BIO®, our engineers were spending around 2 hrs a week filling in timesheets and writing manual claims for allowances, expenses and overtime. By fully automating this process, each engineer saves up to 80 hours per year. Combined with time saved for back-office teams manually inputting and uploading daily work record sheet information equates to around 12,000 hours annually.
Automating admin is a key area that can have a big impact, removing spreadsheet reliance and freeing up people to turn their attentions to more visible and impactful tasks that have a positive influence on customers. When AI works well it should allow ‘people’ to bring more value and creativity to the table.
- Digital Strategy