Singapore Lift Remote Monitoring and Diagnostics: A Practical Guide to AI-Enabled Predictive Maintenance

Understand how sensors, analytics and BCA’s approval pathway can support safer, more predictive lift maintenance.

Lift maintenance is moving beyond a simple calendar-based approach. Connected sensors, remote monitoring and analytics can help building owners and facility teams understand how a lift is performing between scheduled inspections and servicing visits.

However, AI-enabled monitoring is not a replacement for statutory controls, competent lift contractors or human decision-making. In Singapore, owners considering Remote Monitoring and Diagnostics (RM&D) should treat the technology as part of a controlled maintenance system, with clear evidence, governance and approval requirements.

What is lift Remote Monitoring and Diagnostics?

RM&D connects a lift to a monitoring platform that collects operating information from sensors and equipment interfaces. Depending on the lift design and monitoring scope, the data may relate to movement, vibration, temperature, motor behaviour, door operation, ride quality, fault codes, power events or other indicators of equipment condition.

The objective is not simply to collect more data. A useful RM&D system should help teams:

• Detect abnormal operating patterns earlier;
• Identify or narrow down likely fault causes;
• Prioritise issues according to safety, reliability and service impact;
• Support pre-emptive inspection or repair; and
• Create a traceable record of alerts, actions and outcomes.

For a building owner, this can provide better visibility between contractor visits. For a facility manager, it can make fault response more structured. For a lift contractor, it can provide additional condition information before a technician attends site.

Why AI sensors are useful for predictive maintenance

Traditional maintenance schedules are important, but fixed intervals do not always reflect the actual condition or usage of a lift. A lift serving a busy commercial building, logistics facility or residential development may experience changing traffic patterns, loading conditions and environmental exposure.

AI and analytics can compare current readings with historical behaviour or known operating patterns. The system may flag a gradual change that is difficult to identify through occasional observation, such as increasing vibration, longer door cycles or repeated intermittent faults.

This does not mean every alert represents an imminent failure. Alerts should be assessed in context. A practical system distinguishes between a low-priority deviation, a recurring issue that needs planned attention and a condition requiring immediate escalation.

Build the right data foundation

Predictive maintenance is only as reliable as the data and engineering context supporting it. Before installing sensors or selecting software, owners and their appointed contractors should document the lift assets, control systems, existing fault records and maintenance history.

Important questions include:

• Which lift components and operating conditions need to be monitored?
• Can the monitoring system connect safely to existing equipment and building systems?
• Are sensor readings time-stamped, stored and traceable?
• How will missing, noisy or contradictory data be handled?
• Who reviews alerts and decides what action is required?
• How will an alert be linked to a work order, inspection record or closure note?

A sensor dashboard alone is not a maintenance strategy. The data must connect to practical workflows, including escalation, technician review, planned parts replacement, access coordination and post-repair verification.

Use anomaly detection and diagnosis together

Anomaly detection identifies behaviour outside an expected range. For example, a system may notice that a door operation pattern, vibration profile or motor signal has changed over time. This is useful for early warning, but it does not necessarily explain the cause.

Fault diagnosis goes a step further by comparing multiple signals, event histories and operating conditions to help identify possible root causes. The output should be presented as decision support rather than an unquestionable conclusion. A competent person still needs to assess the lift, confirm the condition and determine the safe response.

For facility teams, the most valuable alert is usually not the most technically complex one. It is an alert that states what changed, why it matters, the recommended priority and the next responsible action.

Understand the Singapore approval pathway

BCA’s RM&D guidance describes a pathway for lift owners seeking approval to move towards three-monthly maintenance intervals when the connected monitoring and diagnostic arrangement is suitable. This should not be understood as an automatic right created by installing IoT devices or an AI platform.

The guidance sets out a two-stage process. It includes consultation and a pre-approval trial, followed by a formal application to the Commissioner of Building Control. The trial period is stated as at least six months, allowing the owner and relevant parties to demonstrate that the RM&D system, maintenance process and supporting evidence are working as intended.

Building owners should confirm the current requirements directly with BCA and work closely with their appointed lift contractor and other qualified parties. The approval process may require evidence relating to monitoring coverage, alert handling, system performance, maintenance records and operational controls. The exact submission needs should be checked against the current official guidance and the specific lift installation.

Until the applicable approval is obtained, owners should continue to follow the required maintenance arrangements and statutory obligations. Quarterly maintenance should be treated as an approved, evidence-based operating model, not as a cost-cutting shortcut.

Plan the trial as an engineering programme

A six-month or longer trial should have defined objectives from the beginning. The team should establish a baseline for normal operation, agree what constitutes an alert, record how alerts are reviewed and measure whether the system leads to timely and appropriate actions.

The trial should also test less convenient scenarios: communication loss, sensor failure, false alarms, unusual traffic, power interruptions and faults that do not follow historical patterns. A strong trial demonstrates not only that the dashboard works, but that people can respond safely when the system produces incomplete or unexpected information.

Useful trial records may include alert history, technician findings, corrective actions, unresolved exceptions, system availability information and evidence that maintenance decisions were reviewed by the responsible parties.

Do not overlook cybersecurity and interoperability

Connected lifts create additional digital interfaces that need to be managed. Owners should ask how devices communicate, where data is stored, who can access the system, how accounts are controlled and how software or firmware changes are governed.

Cybersecurity planning should include access control, secure configuration, monitoring of unusual activity, backup arrangements and a process for responding to suspected compromise. It should also clarify responsibilities between the building owner, lift contractor, technology provider, network provider and any facilities management platform.

Interoperability is equally important. A lift monitoring system may need to exchange information with a maintenance management platform, building operations dashboard or enterprise reporting system. Open and well-documented interfaces can reduce manual re-entry, but integration should never bypass safety controls or expose systems unnecessarily.

How ISS can support the implementation

ISS can help organisations assess how sensor data, AI automation and digital workflows could fit into their engineering or facilities management operations. The practical focus should be on connecting information to action: receiving and prioritising alerts, routing tasks, recording decisions, supporting contractor coordination and maintaining an auditable digital trail.

Every lift installation is different. A suitable solution may involve a focused monitoring pilot, integration with existing systems or a broader digital operations workflow. The right starting point is a site and asset review, followed by a clearly defined scope, responsibilities and validation plan.

Final checklist for building owners

• Appoint the appropriate lift contractor and qualified responsible parties early.
• Define the safety and maintenance outcomes the RM&D system must support.
• Map the required sensor data, interfaces and historical records.
• Design alert triage, escalation and work-order workflows before deployment.
• Include cybersecurity, access management and interoperability in the scope.
• Run a controlled trial with documented evidence and human review.
• Check the current BCA guidance and approval requirements before changing maintenance intervals.

Remote monitoring can make lift maintenance more informed and predictive, but its value depends on engineering discipline, reliable data and accountable decision-making. Contact ISS to discuss your engineering, facility management or AI automation requirements.