Key Takeaways:
AI is opening the door to PdM in facilities for managers, staffs and organizations willing to embrace the changes needed to make it happen.
A successful AI-PdM process also requires that managers create a robust path for data to make its way through the various platforms that departments and technicians rely on when planning and prioritizing maintenance activities.
While managers might still hesitate to commit time and energy to tapping into AI to support PdM, there is little doubt that applications of AI in facilities will continue to expand.
For many engineering and maintenance managers, tangible applications and benefits of artificial intelligence (AI) have been rare. But one area in which the rapidly expanding technology shows the most promise — and actually has produced benefits — for departments and operations is the move to predictive maintenance (PdM).
While many departments continue to practice reactive maintenance and some have made the transition to predictive maintenance, relatively few have had the means to implement and practice PdM — until now.
AI is opening the door to PdM in facilities for managers, staffs and organizations willing to embrace the changes needed to make it happen.
Not surprisingly, few managers and their staffs have begun the process of moving their departments’ maintenance strategy from preventive to predictive, much less tapping into the power of AI to help.
“My sense in the industry at large is that most facility managers are still early — somewhere between awareness and experimentation,” says Paul Morgan, chief operating officer, real estate management services and head of workplace management with JLL. “They’ve heard the promise of AI, but relatively few have fully operationalized AI into their workflows.
“The gap isn’t usually interest. It’s trust, data quality and integration. Many organizations have fragmented data across BMS, CMMS, IWMS and spreadsheets, which makes it hard to feed a model reliable input. There’s also a generational split. Newer facility managers tend to be more comfortable experimenting with AI tools, while more tenured managers, who often have the deepest institutional knowledge, are more skeptical until they see the model prove itself over a few real failure events.”
Leobardo Bobadilla is one of the few facilities executives who has made the leap. Bobadilla is vice president of facilities management at Miami Dade College in South Florida. The college has eight campuses throughout the Miami area with about 100 buildings and 8 million square feet of space.
“Our experience with AI started with our academic team on the programmatic side, because the college was the first college in Florida that developed a degree granting program around AI,” he says. To accommodate the teaching and learning the AI program required, Bobadilla and his team needed to figure out the impact on the college’s facilities.
“As we started to do that work, we also started to look at what that means to the facilities management industry,” he says. “That led us to ask our vendors, ‘What were they doing along these lines of AI, and where were they on this learning curve?
“What we found out is that a lot of them were also discovering this space and trying to understand what this meant to their industry. Some of them had pilot programs that were already going on.”
A critical piece of the PdM-AI equation is data — a realization Bobadilla and his team came to soon after their efforts began.
“We have tons of data every day from all the various pieces of equipment that we have operating,” he says. “Historically, most of that data has been looked at when we’re reacting to something. When something is broken down, we’ll go into the system, and we’ll see what the system is telling us. At that point, we’re really trying to figure out how to act on an emergency basis to address the issue.”
“What this is allowing us to do is to really, truly look at a predictive maintenance model where something has not broken down yet. It’s not impacting the customer, but something is not working exactly as it should, and a technician can respond to it.”
Morgan points out three critical reasons managers need to use AI to tap into critical facilities data in their pursuit of PdM:
“Practically, this means starting with high-value, high-failure-cost assets — chillers, rooftop units, elevators — rather than trying to instrument everything at once, as well as using AI recommendations alongside and not instead of technician expertise in the early stages,” Morgan says.
Dan Hounsell is senior editor for the facilities market. He has more than 30 years of experience writing about facilities maintenance, engineering and management.
AI, Predictive Maintenance and the Path to Success
How AI Predictive Maintenance Succeeds: Training, Data Integration and Trust
