How AI, IoT and Best Practices Can Prevent Scaffold & Fire Disasters in Construction Across Asia
On 26 November 2025, in Wang Fuk Court residential complex in Tai Po, Hong Kong, a devastating fire incident showed the underlying risks escalating from a small fire instance.
While the investigation is ongoing, the initial reports suggested the fire that began as a contained incident rapidly escalated along bamboo scaffolding surrounding the site. Within minutes, flames climbed vertically, smoke spread across multiple floors, and emergency responders faced a complex, fast-evolving hazard.
When considering Asia’s dense urban construction environments, scaffolding systems are seen to often coexist with hot works, temporary electrical loads, combustible materials, and narrow access routes. When these elements interact, small deviations can cascade into major incidents. The challenge is no longer a lack of rules or materials knowledge, but the absence of real-time intelligence capable of interpreting how risks combine and evolve.
This is where the role of artificial intelligence (AI), Internet of Things (IoT) sensing, and proven safety practices is reshaping how scaffold and fire risks can be prevented — not just managed.
Why Scaffold-Related Fire Risks Are Systemic in High-Density Construction
As per the U.S. Bureau of Labour Statistics, the construction sector contributes the majority of industrial fatalities. One of the critical areas in the sector includes the use of scaffolds.
Scaffolds in construction sites are often treated as static structures. Those that are erected, inspected, signed off, and assumed safe. But in reality, they are dynamic systems where loads change as materials are stored, the access routes shift, while wind, humidity, and temperature lead to alterations in material behaviour.
In bamboo scaffolds, commonly seen across parts of Asia, drying, cracking, resin exposure, and surface roughness can materially affect ignition potential and flame spread—factors that periodic visual inspections struggle to capture in real time.
Fire risk in such scaffolding environments often extends well beyond these direct ignition sources. Temporary lighting systems around the scaffolds at work can overheat. Improvised electrical connections may degrade under vibration and weather exposure. Protective sheeting can obstruct sprinkler coverage and restrict access for firefighting teams. In high-rise settings, scaffold-wrapped façades allow smoke and heat to bypass internal fire compartments, threatening upper floors before internal alarms fully activate.
In Asia’s high-density cities—where scaffolding often envelops entire buildings—the margin for error is measured in seconds, not minutes.
AI Covering from Visual Inspections to Continuous Risk Awareness
Traditional scaffold safety methods rely heavily on scheduled inspections and manual sign-offs. While necessary, these methods can capture conditions only at specific moments.
But when accompanied by AI-enabled vision systems, this oversight occurs continuously.
For instance, computer vision models trained on construction-specific datasets, when integrated into AI CCTVs on sites or drones, can monitor scaffold alignment, joint integrity, loading patterns, and unauthorized modifications in real time.
A misaligned ledger, excessive material stacking, or removal of toe boards can be detected as soon as they occur and not hours later during a walkthrough.
More importantly, when the use of AI-powered vision analytics is combined with fire-risk context, the system begins to understand interaction, not just structure. For example, identifying hot work activity occurring within unsafe proximity to bamboo scaffolds or detecting sparks in zones where combustible dust or materials have accumulated.
This transforms scaffold monitoring from compliance verification into live risk interpretation.
IoT Sensors Along the Invisible Conditions Driving Fire Escalation
Many of the most dangerous precursors to fires are invisible. When factors like temperature gradients, electrical overloads, humidity shifts, and airflow changes are present, they rarely trigger alarms until thresholds are crossed — often when it is already too late.
IoT sensors embedded across scaffolding zones, electrical panels, and work areas provide the missing environmental layer. Thermal sensors can identify abnormal heat buildup along scaffold surfaces or near temporary wiring. Gas and particulate sensors can detect early smoke particles or off-gassing long before visible flames appear. Vibration sensors can signal instability caused by wind gusts or overloading.
In reference to the deadly Hong Kong fire incident, such data streams could help identify how heat propagated upward along scaffold layers, how airflow accelerated flame spread, and where early intervention may have slowed escalation.
AI as the Interpreter Between Data and Decisions
One thing that needs to be understood here is that data alone does not prevent disasters, but interpretation does. The role of AI in preventing construction fire and scaffold safety lies in its ability to connect patterns across time, space, and behaviour.
By correlating camera feeds, IoT thermal data, electrical loads, weather inputs, and historical incident patterns, the safety models can generate dynamic risk scores for scaffold zones.
A scaffold may be structurally sound in the morning but become high-risk by afternoon due to rising temperatures, increased activity density, and temporary wiring adjustments. The AI-based interventions ensure safety teams receive a timely update on data, rather than the one at the beginning of the task.
Predictive analytics in AI-based systems enable EHS teams to intervene before ignition occurs — by halting hot work, redistributing materials, enhancing fire watch coverage, or adjusting ventilation paths.
As Gary Ng, CEO of viAct, observes,
“Major site incidents rarely begin with a single failure. They emerge when small changes interact faster than humans can process. AI gives safety teams the ability to see how risk is forming — not just when it has already arrived.”
A Turning Point for Construction Safety in Asia
The Hong Kong scaffold fire serves as a stark reminder that in dense, vertical construction environments, risks escalate faster than human observation alone can manage. As Asian countries continue to urbanize at unprecedented speed, scaffold and fire safety cannot rely solely on periodic inspections or reactive alarms.
AI and IoT offer something fundamentally different: it is the continuous awareness, predictive insight, and earlier intervention. When combined with effective engineering practices and disciplined site management, they provide a path toward preventing the first spark — not just responding to the blaze.
Industry Perspective
According to Gary Ng, CEO and Co-Founder of viAct, major site incidents are rarely the result of a single failure but rather the rapid interaction of multiple risk factors.
With a background in building engineering, Gary transitioned into an AI entrepreneur with the founding of viAct in 2016. He brings over a decade of experience in implementing technological innovations across the construction industry, helping organisations move from reactive safety management to predictive, data-driven decision-making.
Prior to viAct, Gary served as Managing Director of 3D fashion technology company EFI Optitex and was recognised as Best Regional Senior Executive within NASDAQ-listed technology enterprise Stratasys. Earlier in his career, he also contributed as an advisory board member for SXSW, reflecting his long-standing involvement in technology-led innovation ecosystems.
Beyond industry leadership, Gary is a strong advocate for knowledge transfer between experienced professionals and the next generation. He is a respected academic contributor and currently serves as a visiting faculty professional at The Hong Kong Polytechnic University. As an active public speaker, he continues to champion AI-driven sustainability and safer workplaces across construction and industrial environments.
For more information please visit viAct website.
