Like every major technological advancement before it, AI is met with trepidation, a tool that some may feel if left unchecked could lead to the ‘rise of the machines’. In the FM sector at least, we are not quite ready for the computers to take over – research by DMA Group in 2021, showed that many FM firms are behind the curve when it comes to embracing technology, with 3 in 10 professionals feeling like their teams weren’t taking advantage of technology.
My feeling is that not much has changed in the last three years. Many organisations are still reluctant to let go of spreadsheets, or modular CAFM systems, so the idea of AI integration may be too much for them to digest at the moment. Is it even possible for AI to be fully integrated into FM as it exists now?
Lessons from other industries
We can look to other sectors to see AI’s current shortcoming. Amazon, for example, had to stop using this technology in recruitment selection when it started blanket-rejecting women. McDonald’s stopped using AI-generated drive-throughs after, among other errors, it tried to add 260 items to a customer’s order. In one of its most serious error, AI has recommended unsafe and incorrect forms of cancer treatment.
While AI is continuously evolving it’s important to remember that it is a very new technology and mistakes can happen. We cannot yet (if ever) rely entirely on AI; human input is almost always essential but isn’t infallible either.
AI pitfalls in FM:
In the FM sector, there is still much to be done before the potential of AI can be truly realised:
Missing data that promotes false machine learning: AI is only able to work based on its presented data. While this might be imputed automatically, what does this produce? Averaging using the mean, median and mode might result in biased results. After all, machines are not equipped to handle nuance.
The data might be missing altogether, due to human error, inaccessibility, because of malfunctions at the data collection sensors or a network issue. With incorrect data, the results can be disastrous. Energy usage, for example can be skewed with incorrect data, adding to increased costs and the potential of moving away from sustainability goals.
Contextualising information presented by AI: As FM is an ever-changing world, we must ask ourselves whether the data can produce accurate machine learning. We must consider the sample bias that could throw everything off.
Think about occupancy data. This is used to detect what rooms are used within a building and when, which can allow for resource allocation and asset repair. AI-integrated HVAC that uses inaccurate occupancy data, for example, might turn up the heating incorrectly, leading to uncomfortable occupants and unnecessary cost.
Ignoring the importance of ‘humanity’: The example from Amazon shows that there are some jobs that hinge on the human touch. Recruiting the right person should be based on more than box ticking, especially if that machine has inadvertently learnt to be sexist! In FM, customer relations are key, and maintaining and developing these relationships relies heavily on human interaction.
A 2024 global consumer trends report from qualtrics found that most people prefer to engage with a human at some point in customer service. Apart from being part of our biological make-up as social animals, AI currently struggles with creative problem solving that might appear out of the box, something that humans excel in by comparison (at the moment!)
FMs on the frontline have to be agile and flexible in their thinking, adapting to changing project demands and unforeseen problems, all while being sensitive to the needs of the customer. A database, all be it a state of the art continually learning one, cannot replicate these soft skills.
Ethical considerations: One of the main concerns about AI usage is about the protection and privacy of data. When AI collects data, it begs the question who ends up owning the data and how it is used? Clear guidelines must be put in place of how information can be accessed and stored.
The cost of AI: Like any new technology, integrating AI within your FM business will cost money. There should be savings made by taking a digital approach – improved working practices, energy savings and enhanced customer satisfaction, for example. Your customer may not be willing to share this investment, even though they will no doubt reap the benefits down the line, so you must consider the ROI and whether your business is in a position to shoulder the initial outlay.
A supporting role
There is no doubt that AI has the potential to be transformative, but for most FM businesses there is a lot to be done before its benefits can be realised. Poorly implemented AI, where data volume and quality is
insufficient, for example, will only lead to problems, which in our sector could be very costly mistakes. At the moment, at least, AI must take a supporting role for all but the most agile and already tech-savvy firms.