Editor’s note: FM Perspectives are industry op-eds. The views expressed are the authors’ and do not necessarily reflect those of Facilities Management Advisor.
For decades, facilities management leaders have attempted to bridge the gap between preventative and reactive maintenance—now, the rapid onset of new technologies gives providers vastly improved methods of detecting when systems may be at risk. But have these methods become infallible, or is there a way to go before the FM industry shifts to complete reliance on predictive maintenance?
The Rise of Predictive Tech
Since the mid-20th century, specialists within certain sectors have been able to manually conduct tests to predict when systems were failing or degrading, making it easier for work to be planned in advance—meaning a reduction in downtime and fewer unexpected costs. As maintenance management systems became computerized at the turn of the millennium, this approach became less time-consuming and specialized, opening up the opportunities for a broader range of industries to utilize a predictive approach.
Yet it took the invention and implementation of IoT sensors, AI technologies, and real-time data availability for predictive maintenance to become truly accessible, enabling FM providers to enhance its use within their day-to-day operations and to provide a firmer guarantee of its accuracy to their clients. Technology embedded into building maintenance and building automation systems, and the widespread implementation of cloud platforms and data process systems, have reduced some of the workload involved in interpreting information and sensor readings, and we are now in the era of automated data analysis. This evolution provides a host of benefits but also raises questions about how much we can (and should) rely on technology, especially given the high costs of doing so, and the even higher costs if this approach turns out to be the wrong one.
A Happy Medium?
The attraction of predictive maintenance is clear, especially in a world where our buildings rely on more complex systems than ever to ensure safety, comfort, and ease of use—from the basics like temperature and lighting, to the smart tech used for access and security. In theory, if we can plan when work will be required, those using the building can be notified ahead of time, schedules can be set to cause the least disruption, and any replacement parts can be ordered in advance to minimize downtime.
This approach also avoids the sometimes unnecessary costs associated with preventative maintenance; whilst in theory, having a set calendar should keep systems in good working order, it means undertaking that task because it is in the diary—not because it is absolutely necessary at that point in time.
For FM teams, predictive maintenance allows them to allocate labor and contractor visits more efficiently, avoiding both unnecessary routine interventions and unplanned failures. In fact, Deloitte found on average there were 70% fewer breakdowns, with maintenance costs being 25% lower and maintenance planning time reduced by up to half.
The Future of FM?
However, predictive maintenance is not a flawless method—and there are drawbacks to consider. As with any approach which relies on technology, it can only be as good as the data underpinning it. Without well-connected assets, integration across systems, and good-quality, properly analyzed data, the approach becomes dramatically less reliable.
Particularly when it comes to older buildings, switching to a predictive way of working may involve initial investment to upgrade legacy systems and to ensure all of the technology can communicate effectively. There may also be associated training costs, with FM teams needing to be able to interpret data and competently use new digital tools to ensure success.
As well as anecdotally, these concerns have been highlighted through research: A systematic review found insufficient data was the “number one barrier” to full AI implementation, while studies from both CBRE and JLL have cited data quality as a vital component of success, with the latter also focusing on the need to address legacy system integration.
All of these issues can be overcome, of course, but they do require a level of time and cost investment which may outweigh the benefits—meaning the industry is seeing predictive maintenance being utilized for critical assets, larger portfolios, and smart buildings, but providers are still relying on preventative maintenance in other situations (and resorting to reactive maintenance where a problem has not been highlighted in advance of system failure).
Given the huge advances in the amount and accuracy of data we can now gather, it may be that a decade down the line, it is truly possible to rely fully on predictive maintenance. By then, more of the legacy systems in place now will have been replaced with modern alternatives, and far more specialists will have experience in analyzing predictive data. But far more work is needed—both by individual providers and the wider built environment industry—before we can truly say that predictive maintenance is the future of FM.
Adam Atkins is the co-founder and chief executive of Coat Facilities Group.