From cleaning floors to driving portfolio value: How AI will disrupt facility management

Artificial intelligence may prove the decisive factor in nudging building owners in Singapore to finally embrace outcome-based contracts

For years, the shift from task-based contracts to performance-linked models has been held back in facility management (FM) by one persistent problem: how to measure outcomes fairly.

What counts as an “urgent” incident varies across industries, thresholds for downtime differ, and customer satisfaction is hard to quantify. Building owners worry that outcomes cannot be measured consistently, while operators struggle with fragmented data streams that are difficult to synthesise into clear actions for asset owners.

The result is that most contracts still resemble old-style service agreements, pegged to manpower counts and prescriptive task lists.

Service providers are paid for daily cleaning, monthly air-conditioning filter changes, or ensuring 95 per cent of lights are working. Compliance is typically proven through logs, reports, and inspections – processes that look rigorous but often prevent efficiencies, such as reducing cleaning in unused spaces or scaling back cooling on school campuses during term breaks.

The advent of sensors and Internet of Things (IoT) devices has accelerated the pace of automation in the labour-intensive security and facility management industry.

This has been a necessary evolution. Costs of security and cleaning services are rising, squeezing margins for service providers and delivering lower value for building owners.

SJ Group estimates that Singapore’s progressive wage model roll-out could push these costs up by 15 to 20 per cent in the next three years by 2028. Coupled with tighter climate regulatory standards and increasing demand for high performing investor-grade assets, the pressure to optimise is immense.

Artificial intelligence (AI) offers the next big driver to create more effective ways to manage and operate facilities, both for service providers having to navigate outsourcing costs, and for building owners to keep operational expenses in check.

From sensors to predictive intelligence

Stadiums, airports, hospitals and campuses have deployed sensors and IoT over the last decade to move from a time-based, task-based contracting model to one that guarantees equipment uptime or energy savings. Yet, the model remains largely reactive; maintenance tickets are triggered when a threshold is breached, often after deterioration already sets in.

Today, AI promises to transform this landscape. With years of accumulated sensor data and maintenance logs, the building management sector is anticipating a future where predictive models can now forecast failures before they occur. By analysing vibration, temperature, pressure, humidity, and thermal imaging data, AI can estimate the remaining useful life of assets and flag risks well in advance.

This brand of predictive maintenance allows contracts to move beyond traditional service levels. Instead of stipulating “repair faulty air-conditioning within two hours”, contracts can guarantee performance outcomes, such as limiting unplanned breakdowns to less than 2 per cent annually.

As models mature, AI will also suggest optimal energy settings, adjust comfort levels automatically, and even plan cleaning schedules based on occupancy.

These outcomes align with broader business goals for building owners, which include meeting climate regulatory standards through carbon reductions, energy efficiency gains and ensuring that buildings continue to attract investor interest. High-performing buildings are usually characterised by healthy occupancy levels and measured by positive tenant ratings for safe, clean spaces that promote positive well-being.

AI as a strategic asset tool

The real transformation lies in how AI turns performance-based contracting from an operational efficiency play into a strategic lever for long-term asset planning. By continuously collecting and analysing data on asset performance, from energy use and equipment reliability to occupancy trends, AI generates insights that go well beyond daily maintenance.

These insights enable owners to make evidence-based decisions about the next iteration of their assets. These could take the form of adaptive re-use strategies, asset enhancement initiatives, and the creation of new revenue streams.

Upgrading core systems such as air conditioning, lighting, and facades not only improves efficiency and user comfort but can also position a property for premium rents, while smart building upgrades and sustainability retrofits boost long-term value and appeal to sustainability-focused investors.

At the same time, owners can monetise underutilised areas by introducing rooftop solar, electric-vehicle charging bays, digital advertising, or amenity-driven offerings such as food halls, wellness spaces, and flexible leasing models. Together, these strategies generate fresh income opportunities while strengthening tenant retention and overall portfolio performance.

For instance, predictive models might reveal that a chiller system, though serviceable, will cost more to maintain than to replace within five years, prompting a capital investment that pays off in lower energy bills and reduced carbon output. Occupancy analytics may show persistent underuse of a campus block, strengthening the case for adaptive re-use into co-working, residential, or community functions.

In this way, AI-powered outcome-based contracts create a continuous feedback loop between operations and design, helping asset owners future-proof portfolios, optimise lifecycle costs, and align buildings more closely with evolving business needs and sustainability goals.

There is, however, no ignoring the upfront costs. Deploying smart facility management tools requires significant investment in hardware, software, and integration.

But instead of treating these as sunk costs, owners and operators can structure outcome-based contracts so that efficiency gains and cost savings fund the technology. Lower energy bills, reduced downtime, and extended asset life can all be shared transparently between both sides.

This win-win model aligns incentives and ensures both owner and service provider are rewarded for long-term performance, not just compliance.

Elevating the workforce

Education and training pathways must evolve in parallel.

For decades, FM has been a dependable career track for polytechnic and university graduates in the built environment. Curricula will now need to expand beyond the nuts and bolts of traditional facility management operations and workplace safety to include analytics, data literacy, AI platforms and digital twin (virtual replicas of building systems) management.

Engineering remains a core discipline, providing the grounding to ensure AI predictions are interpreted correctly.

In this new era, facilities shift from cost centres to strategic assets, future-proofed against rising costs, regulatory demands, and the evolving expectations of tenants and investors alike.

The writer is senior director, facilities and asset management at SJ Group

Published by The Business Times, 9 September 2025