Singapore Alternative Cooling Technologies: Using AI Monitoring to Optimise ACMV Energy and Comfort

A practical guide to selecting, commissioning and monitoring alternative cooling systems in Singapore buildings and warehouses.

Singapore businesses are under increasing pressure to improve energy performance without compromising occupant comfort, indoor air quality or operational reliability. For many buildings and warehouses, the answer may not be a simple replacement of one air-conditioning system with another. Alternative cooling technologies, supported by sensors and intelligent controls, can provide a more targeted route to reducing cooling demand.

Singapore’s Building and Construction Authority has highlighted alternative cooling technologies, or ACTs, as part of the built environment’s wider energy-efficiency and decarbonisation pathway. Examples include hybrid cooling approaches and passive displacement cooling. However, the technology itself is only one part of the business case. Owners and facility teams also need to confirm that the system performs properly in Singapore’s hot and humid conditions.

This is where AI monitoring becomes useful. Rather than treating AI as a standalone energy product, businesses can use it as a measurement, diagnostics and optimisation layer across the ACMV system.

What are alternative cooling technologies?

Alternative cooling technologies are approaches that reduce reliance on conventional mechanical cooling or use it more efficiently. The appropriate solution depends on the building’s layout, occupancy, ventilation requirements, operating hours and internal heat loads.

Hybrid cooling combines different cooling or ventilation methods so that mechanical cooling is used only when required. In suitable conditions, this can reduce the operating hours or load of conventional ACMV equipment. The system must still be designed and controlled carefully because outdoor humidity, temperature and indoor air-quality requirements affect when different modes are appropriate.

Passive displacement cooling uses the movement and stratification of air to deliver cooler air where it is needed and allow warmer air to rise away from the occupied zone. It may be relevant in selected commercial, industrial or high-volume spaces, but performance depends on ceiling height, air distribution, heat sources, occupancy and space usage.

Other solutions may also be considered depending on the project. BCA’s Alternative Cooling Technologies resources provide technology categories, use cases and indicative information on energy savings, cost and payback. These figures should be treated as guidance rather than guaranteed results because actual performance is project-specific.

Why monitoring matters after installation

A cooling project can appear successful during handover but perform differently during daily operations. Controls may be overridden, sensors may drift, filters may become blocked, doors may remain open, or occupancy patterns may change. A warehouse may also have different conditions between loading bays, storage areas, offices and temperature-sensitive zones.

Post-installation monitoring helps answer practical questions:

• Is the occupied zone within the required comfort range?
• Is indoor air quality being maintained during peak occupancy or loading activity?
• Is the alternative cooling mode operating when conditions are suitable?
• Are fans, pumps, dampers and chillers working harder than expected?
• Is energy use improving compared with the agreed baseline?
• Are repeated alarms pointing to a maintenance or control problem?

Without this information, a facility team may not know whether higher energy consumption is caused by weather, occupancy, equipment degradation, poor scheduling or an incorrect control sequence.

What should an AI monitoring layer measure?

The monitoring plan should be based on the building’s operational risks and performance objectives. A typical system may combine data from:

• Temperature and relative humidity sensors in representative occupied areas
• Carbon dioxide or other indoor-air-quality indicators where relevant
• Airflow, pressure and differential-pressure measurements
• Chilled-water temperatures, flow rates and equipment status
• Fan, pump and compressor operating data
• Electrical submeters for cooling equipment and major loads
• Occupancy, operating schedules and access or booking data where available
• Outdoor weather information and relevant environmental conditions

Data quality is critical. Sensors should be located appropriately, checked during commissioning and maintained over time. A sophisticated dashboard cannot compensate for missing, poorly positioned or unreliable inputs.

How AI can support facility teams

AI and analytics are most useful when they help people make better operational decisions. For example, a monitoring platform can establish normal operating patterns and flag unusual behaviour, such as a fan running outside scheduled hours, a temperature zone drifting gradually, or cooling energy rising while occupancy remains stable.

AI-assisted fault detection can also help prioritise maintenance. Instead of sending technicians to investigate every alarm, the system can group related symptoms and identify likely causes for review. This does not replace engineering judgement or physical inspection. It helps the facility team focus attention where it is most needed.

For automated control, the system may adjust schedules, setpoints or equipment staging within approved operating limits. In a Singapore facility, controls should account for humidity, ventilation demand, business-critical areas and the need to avoid short cycling or excessive equipment switching. Any automation should include clear override functions, alarm escalation and a safe fallback mode.

A practical implementation approach

1. Establish a baseline

Review historical energy data, operating schedules, temperature complaints, maintenance records and current ACMV performance. Where possible, separate cooling energy from other building loads. A baseline should reflect actual operating conditions rather than an assumed design profile.

2. Map zones and use cases

Divide the building into meaningful zones. An office, server room, warehouse picking area and loading bay will not have the same comfort, ventilation or operating requirements. Identify where alternative cooling is technically suitable and where conventional cooling remains necessary.

3. Define acceptance criteria

Agree on what success means before installation. This may include energy performance, temperature and humidity stability, indoor-air-quality indicators, response time, equipment runtime and maintenance outcomes. Avoid relying on a single energy figure.

4. Integrate data carefully

Connect relevant meters, sensors, building-management systems and equipment controllers. Standardise naming, timestamps, units and alarm priorities. Data ownership, cybersecurity, user permissions and system resilience should also be addressed during design.

5. Commission in operating conditions

Test the system during representative periods, including different occupancy levels and operating modes. Confirm sensor readings, control sequences, alarm handling and manual overrides. For warehouses, include practical scenarios such as loading activity, door opening and changing shift patterns.

6. Review performance continuously

Use weekly or monthly reviews to compare actual performance with the baseline and acceptance criteria. Investigate exceptions rather than reacting only to comfort complaints. The monitoring system should produce clear actions for the facility team, not just charts.

Common mistakes to avoid

• Choosing technology before assessing the space: A solution suitable for one building may not suit another.
• Measuring only total electricity: Whole-site consumption can hide whether cooling performance improved.
• Ignoring humidity and ventilation: Lower temperature does not automatically mean better comfort or indoor air quality.
• Automating poor operating practices: AI cannot fix incorrect schedules, leaking doors, blocked filters or badly configured equipment without the right operational response.
• Installing too few sensors: A single reading may not represent a large warehouse or multi-zone building.
• Failing to plan for maintenance: Sensors, meters and controllers require calibration, support and clear ownership.

Planning the business case in Singapore

When comparing solutions, consider capital cost, installation disruption, maintenance requirements, expected energy performance, comfort risk and the ability to expand the system later. BCA’s updated ACT resources can support early evaluation, while Enterprise Singapore’s Energy Efficiency Grant may be relevant for eligible businesses and equipment investments. Eligibility, support levels and application conditions should be checked against the latest official requirements before making financial assumptions.

Singapore’s wider built-environment strategy is moving towards measurable energy efficiency and operational performance. For owners, facility managers and warehouse operators, this creates an opportunity to combine engineering improvements with practical digital monitoring.

Alternative cooling should not be treated as a one-time equipment purchase. It is an operational system that needs commissioning, data, feedback and continuous improvement. AI monitoring can help businesses verify whether the technology is delivering the intended balance of energy efficiency, comfort, indoor air quality and reliability.

ISS can help businesses discuss the engineering, facility-management and AI automation requirements behind an ACMV optimisation programme. Contact ISS to explore a practical approach for your building, warehouse or SME facility.