Construction Safety in Singapore: WSH Regulations, Hazards & AI Solutions
“Quick AI-Powered Insights on the Topic— Freshly Updated!”
|
Construction safety in Singapore is governed by the Workplace Safety and Health (WSH) Act and its regulations, which require contractors to find hazards, control the risks and keep records. Since 1 June 2024, every construction site with a contract value of S$5 million or more must install a video surveillance system (VSS), and the maximum fine for serious breaches has risen from S$20,000 to S$50,000.
This guide explains the main WSH rules for construction sites, the most common hazards, what contractors must do, and how technology such as AI video analytics can help meet the requirements.
What Is Construction Safety in Singapore?
Construction safety in Singapore means protecting workers and others from harm on building and infrastructure sites by following the WSH Act and its regulations. The Ministry of Manpower (MOM) sets and enforces these rules, and the Workplace Safety and Health Council supports industry with guidance.
In practice, construction safety covers three things:
Preventing hazards: guardrails, safe access, controlled lifting and safe vehicle movement.
Managing risk: assessing each job before work starts and reviewing the controls regularly.
Proving compliance: keeping records, inspections and, on larger sites, video footage.
Under MOM's WSH Act, the aim is the safety, health and welfare of everyone at work. viAct supports this on site with a construction AI safety solution that works on existing cameras.
Why Is Construction Safety Important in Singapore?
Construction safety is important in Singapore because it helps prevent workplace fatalities and serious injuries, protects workers from high-risk activities, and reduces project disruptions caused by accidents, investigations and work stoppages. For contractors, maintaining safe worksites is also essential for meeting Workplace Safety and Health (WSH) requirements and ensuring that construction activities continue without unnecessary delays.
Despite improvements in workplace safety, construction remains one of Singapore's higher-risk sectors. According to MOM's Workplace Safety and Health Report 2025, the sector's fatal and major injury rate declined from 31.0 per 100,000 workers in 2024 to 26.3 in 2025.
Key findings from MOM include:
Small-scale construction risks: Addition and alteration works, renovations and other smaller-scale projects accounted for more than 60% of construction fatal and major injuries in 2025.
High-risk industries: Construction and manufacturing together accounted for 44% of fatal and major injuries in 2025, according to MOM's March 2026 speech.
Recent safety performance: Construction recorded 69 fatal and major injuries in the first half of 2026, with an annualised rate of 24.1 and the lowest since 2020. However, smaller-scale construction works experienced a slight increase.
National safety target: Singapore aims to reduce its national workplace fatal injury rate to below 1 per 100,000 workers by 2028.
Beyond preventing injuries, effective construction safety management is also required to help minimise work stoppages, avoid costly rework and maintain project schedules. This makes construction safety both a regulatory responsibility and an important part of reliable project delivery.
What Are the Main WSH Regulations for Construction Sites in Singapore?
In Singapore, construction safety is primarily regulated under the Workplace Safety and Health Act and its subsidiary legislation. The Ministry of Manpower (MOM) enforces these requirements, while the Workplace Safety and Health Council (WSHC) provides industry guidance and recommended practices.
The following regulations are particularly relevant to construction contractors.
Regulation | Main requirements for contractors |
Workplace Safety and Health Act | Establishes workplace safety duties for employers, occupiers, principals and other responsible parties |
WSH (Construction) Regulations | Covers safety requirements for construction activities, worksites, equipment and work processes |
WSH (Risk Management) Regulations | Requires workplace risk assessments and implementation of appropriate risk controls |
WSH (Work at Heights) Regulations | Addresses fall prevention, safe access and requirements for hazardous work at height |
WSH (Operation of Cranes) Regulations | Establishes requirements for safe crane operations and lifting activities |
WSH (Scaffolds) Regulations | Covers scaffold design, erection, use, inspection and maintenance |
WSH (Safety and Health Management System and Auditing) Regulations | Establishes management system and audit or review requirements for applicable workplaces |
Meeting these requirements begins with understanding the hazards present on construction sites and applying suitable preventive measures.
What Are the Most Common Construction Site Safety Hazards in Singapore?
Construction hazards arise from the activities, equipment and environmental conditions present at a worksite. Some hazards remain throughout a project, while others change as construction progresses.
Understanding these risks is the first step towards selecting effective safety measures.
Falls from heights are a major workplace safety concern in Singapore, particularly during roofing, scaffolding and elevated construction activities. The WSH Council identifies falls as a leading contributor to workplace deaths and highlights the importance of protecting workers near open edges and floor openings.
Workers may fall from scaffolds, ladders, roofs or unprotected building edges due to missing guardrails, unsecured openings or improper use of fall protection equipment. Contractors should provide suitable working platforms, edge protection, safe access and appropriate fall protection systems.
Falling objects remain a serious construction hazard. In January 2025, the WSH Council issued an industry advisory following several fatal incidents involving workers struck by falling objects, including incidents associated with lifting, formwork and demolition activities.
Common risks include improper rigging, unstable loads, equipment failure and workers entering lifting exclusion zones. Contractors should establish lifting plans, use suitable equipment and ensure that trained personnel supervise lifting operations.
Moving vehicles and heavy machinery present serious collision risks, particularly where excavators, forklifts and trucks operate near workers. According to MOM's September 2026 WSH report, vehicular incidents were the leading cause of workplace fatalities across Singapore in the first half of 2026, accounting for 9 of 21 deaths.
Restricted visibility, reversing movements and inadequate separation between vehicles and pedestrians increase collision risks. Contractors should establish designated traffic routes, pedestrian walkways, appropriate warning systems and suitable vehicle safety controls.
Unsafe scaffolds and temporary structures can collapse or cause workers to fall when they are incorrectly assembled, overloaded or modified without proper supervision. In a published enforcement case, the WSH Council documented a worker who fell 3.5 metres from a mobile tower scaffold that lacked proper guardrails and toe-boards.
Contractors should ensure scaffolds are erected and inspected by competent personnel, maintained in safe condition and used within their intended loading limits.
Excavation work exposes workers to collapsing soil, unstable retaining walls, underground utilities and unsafe access conditions. In September 2025, a worker died at a Singapore construction site when an unsupported brick retaining wall collapsed during excavation of a 0.6-metre-deep trench, according to a WSH Council advisory published in March 2026.
Contractors should assess ground stability, provide suitable shoring or other support, locate underground services and inspect excavations before workers enter them.
While these construction hazards require physical safety controls, inspections and worker supervision, unsafe activities can develop between scheduled checks. AI video analytics helps address this monitoring gap by analysing CCTV footage to identify selected hazards and provide contractors with an additional way to detect risks earlier and support timely corrective action.
How Can AI Video Analytics Improve Construction Safety in Singapore?
AI video analytics improves construction safety by analysing camera footage to identify selected unsafe activities, generate alerts and provide records for safety inspections and investigations. It can support contractors in monitoring high-risk activities across large or complex worksites.
Unlike conventional CCTV systems, which often require personnel to review recordings manually, AI-powered systems can detect predefined safety events and notify supervisors when intervention may be required. For example, if a worker enters a designated lifting exclusion zone, the system can detect the intrusion and notify the responsible safety personnel, allowing them to assess the situation and intervene.
Key Applications of AI Video Analytics in Singapore Construction Safety
AI Application | Construction Safety Use Case |
PPE Compliance Enforcement | Identifies workers without required visible PPE, such as safety helmets or reflective vests. |
Fall Protection Monitoring | Detects workers entering defined hazardous areas near open edges and elevated work zones. |
Lifting Exclusion Zone Monitoring | Identifies unauthorised entry into designated lifting zones. |
Vehicle–Pedestrian Detection | Monitors selected unsafe interactions between workers and moving vehicles. |
Restricted Zone Intrusion Detection | Detects people entering hazardous or unauthorised work areas. |
Scaffolding & Temporary Structure Monitoring | Detects unsafe scaffolding conditions, missing safety barriers and potential hazards around temporary structures. |
Incident Documentation | Creates time-stamped event records for review, reporting and corrective actions. |
How Does AI Video Analytics Work With Existing CCTV Cameras on Singapore Construction Sites?
AI video analytics can work with compatible existing CCTV cameras by processing their video feeds through computer vision software. This allows contractors to add automated safety monitoring capabilities without necessarily replacing their entire camera infrastructure.
A typical AI video analytics system follows five steps:
Step 1: Video Capture
Existing CCTV or IP cameras capture activities across construction areas such as lifting zones, access routes and elevated work platforms.
Step 2: AI Video Processing
Video feeds are processed through compatible edge devices, on-premises servers or cloud-based infrastructure, depending on the deployment architecture.
Step 3: Hazard Identification
Computer vision models analyse the footage to identify predefined events, such as missing visible PPE or entry into restricted zones.
Step 4: Safety Alerts
When a configured event is detected, the system sends an alert to designated safety personnel through the supported notification channels.
Step 5: Event Reporting
Detected events are recorded with timestamps and relevant footage, allowing safety teams to review incidents, identify recurring risks and track corrective actions.
A total of 7,000 working hours were saved by a Singapore construction giant between May 2025 and April 2026 when viAct centralised platform, viHUB, helped the EHS teams to generate automated trend reports on recorded PPE violations, heavy machinery near-misses, and hazards involving open edges and confined spaces.
This demonstrates how automated event reporting can turn recorded safety observations into actionable insights, helping EHS teams identify recurring hazards, prioritise corrective actions and reduce time spent on manual safety documentation.
What Are the Key Benefits of AI Video Analytics for Singapore Construction Contractors?
The main benefits of AI video analytics in construction are earlier hazard identification, reduced manual monitoring, more consistent incident documentation and improved visibility across work zones.
Unlike the applications discussed earlier, these benefits concern how AI-assisted monitoring can improve everyday safety operations.
1. Earlier Identification of Unsafe Activities
AI video analytics can detect predefined unsafe events as they occur, reducing the delay between an observable hazard and its identification. For example, an alert involving unauthorised entry into a lifting zone can help a supervisor assess the situation before work continues.
A useful performance indicator is the time between event detection, alert delivery and supervisory response.
2. Reduced Manual CCTV Review
Reviewing footage from multiple cameras can require considerable time, particularly on large construction sites. AI video analytics can flag selected events automatically, allowing safety personnel to focus on relevant footage rather than continuously reviewing every camera feed.
Contractors can assess this benefit by measuring the time spent reviewing footage, preparing incident reports and locating relevant video evidence.
3. More Consistent Incident Documentation
AI systems can create time-stamped records of detected events, including the camera location, event category and relevant video footage. These records can support investigations, safety meetings and corrective action reviews.
The benefit is not simply faster reporting. Consistent records can also make it easier to identify recurring problems that may otherwise remain undocumented.
4. Improved Visibility Across Multiple Work Zones
Large construction projects often involve simultaneous activities across different floors, access routes and work areas. A centralised AI monitoring platform can consolidate selected safety events from multiple cameras, allowing EHS teams to review risks across monitored locations.
This is particularly useful when safety personnel cannot physically observe every activity at the same time.
5. Better Identification of Recurring Safety Risks
Individual safety alerts can reveal immediate problems, while repeated alerts may indicate wider operational weaknesses. For example, recurring vehicle–pedestrian proximity events at the same crossing may suggest a need to redesign traffic routes or improve physical separation.
Contractors can use event frequency, location patterns, response times and corrective action closure rates to evaluate whether safety measures are working.
A 50% reduction in Total Recordable Incident Rate (TRIR) was reported across viAct construction deployments during FY2024–FY2025, covering more than 32,000 workers across 400-plus construction sites. The company-reported results illustrate the potential contribution of AI-based safety monitoring alongside existing workplace controls. They represent global construction deployments rather than Singapore-only results and should not be interpreted as a guaranteed outcome for every project.
The value of AI video analytics therefore depends on how effectively detected events are translated into preventive action, rather than the number of alerts generated alone.
How Can Singapore Contractors Implement AI Video Analytics on Construction Sites?
Implementing AI video analytics requires contractors to identify high-risk activities, assess existing camera infrastructure, configure suitable detection capabilities and establish procedures for responding to alerts.
A practical implementation process includes five steps.
Implementation Step | Key Action |
1. Identify Safety Priorities | Select hazards such as PPE violations, lifting zone intrusions and vehicle–pedestrian interactions. |
2. Assess Existing CCTV | Check camera coverage, image quality, connectivity and compatibility. |
3. Configure AI Detection | Select relevant detection models and define monitored areas and alert conditions. |
4. Establish Alert Procedures | Assign personnel to review notifications and initiate corrective actions. |
5. Evaluate Performance | Review detection accuracy, recurring incidents, response times and system effectiveness. |
Contractors should begin with clearly defined monitoring objectives rather than attempting to automate every safety activity simultaneously.
For example, a construction site experiencing frequent vehicle–pedestrian near-misses may prioritise monitoring around vehicle routes and crossing points. A high-rise project may instead focus on lifting exclusion zones and hazardous access areas.
Once the initial monitoring system is operating reliably, contractors can expand coverage to other activities based on site-specific risks.
Conclusion: Key Takeaways
Singapore contractors must comply with applicable WSH regulations, conduct risk assessments and implement appropriate preventive measures to maintain safe construction sites.
Falls from heights, lifting operations, vehicle–pedestrian collisions and unsafe scaffolding remain major construction hazards that require continuous attention and effective safety controls.
AI video analytics helps identify selected unsafe activities in real time, including PPE violations, lifting exclusion zone intrusions and unsafe vehicle–pedestrian interactions.
Integrating AI video analytics with compatible existing CCTV infrastructure enables automated hazard detection, timely safety alerts and digital incident reporting without necessarily replacing existing cameras.
AI-powered safety monitoring can reduce manual surveillance efforts, improve visibility across construction sites and support data-driven corrective actions when combined with established safety procedures.
The future of construction safety in Singapore lies in moving from reactive incident management to proactive, data-driven risk prevention, where intelligent monitoring helps make safer worksites a continuous operational priority.
Quick FAQs
1. Is AI Mandatory for Construction Safety in Singapore?
AI video analytics is not universally mandatory for construction sites in Singapore. However, since 1 June 2024, construction worksites with contract values of S$5 million and above must install Video Surveillance Systems (VSS) under MOM requirements. Contractors can integrate AI video analytics with compatible VSS infrastructure to automate selected hazard detection, monitor unsafe activities and support safety reporting.
2. Can AI Video Analytics Work With Existing CCTV Cameras on Construction Sites?
Yes, AI video analytics can work with compatible existing CCTV or IP cameras without requiring a complete surveillance system replacement. Cameras generally need to support accessible video streams, such as RTSP, and provide sufficient resolution and visibility for the intended detection tasks. Compatibility also depends on network connectivity, camera positioning and the AI system's processing requirements.
3. How Accurate is it in Detecting Construction Safety Hazards?
For AI platforms like viAct accuracy depends on camera quality, lighting, site conditions and the specific hazard being monitored. Detection performance can vary between applications such as PPE monitoring, lifting exclusion zone detection and vehicle–pedestrian monitoring. Contractors should evaluate detection accuracy, false alerts and missed events through a site-specific pilot before full deployment. Configurable detection thresholds and alert rules can also help reduce unnecessary notifications.
4. Can AI Monitoring Run On-Premises Without Sending CCTV Footage to the Cloud?
Yes, it can operate through on-premises, edge-based, cloud or hybrid deployment models, depending on the provider. On-premises and edge deployments can process footage locally, reducing dependence on continuous internet connectivity and external video transmission. Hybrid systems may process footage locally while sharing selected alerts or incident clips with a central platform. Contractors should confirm data storage locations, access controls and network requirements before implementation.
5. How Much Does AI Video Analytics Cost for a Construction Site in Singapore?
The cost of AI video analytics in Singapore depends on the number of cameras, safety detection modules, deployment model, hardware requirements and system integrations. Pricing may involve per-camera or per-module subscriptions, licensing fees and installation or maintenance costs. Contractors should request a site-specific quotation and consider a proof of concept (POC) to evaluate detection performance, false alerts, operational benefits and total cost before committing to a full deployment.
Read More:
