Project Summary
NovaPeak is supporting a large-scale façade audit inspection programme for HDB residential properties in Singapore, using drone technology, AI-assisted defect analysis and structured engineering reporting.
The publicly announced programme covers façade inspection of more than 1,400 HDB residential blocks, alongside a separate Periodic Façade Inspection programme for 50 multi-storey car parks.
Unlike a single-building inspection, a programme of this scale requires much more than the ability to fly drones. It requires coordinated deployment across hundreds of locations, management of aviation permissions, consistent image capture, defect classification, privacy controls, engineering review, reporting and continuous progress management.
The programme therefore represents a shift from individual drone missions toward industrialised building inspection at portfolio scale.
The Challenge: Inspection at Scale
Inspecting one building can be managed as a project.
Inspecting more than a thousand buildings requires a system.
Every residential block presents different conditions: height, geometry, surrounding roads, pedestrian activity, nearby structures, airspace restrictions and available launch areas.
The inspection programme also has to maintain consistency. Images captured by different teams on different days must still provide sufficient coverage and traceability for subsequent assessment.
HDB’s technical requirements specify 100% visual inspection of façade features, including roofs, together with identification and classification of façade defects.
The challenge was therefore to create a workflow capable of repeating this process reliably across a very large portfolio.
A Multi-Method Drone Inspection Strategy
Not every HDB block can be inspected using exactly the same aircraft configuration.
The programme therefore provides for several inspection methods according to site and aviation constraints.
Conventional drone systems are used where normal flight operations are appropriate. Power-tethered drone systems provide another option for taller buildings, restricted airspace or locations specified by the aviation authority. Where drone flight cannot be carried out, the programme also provides for routine visual façade inspection using an alternative method.
This is important operationally.
The objective is not to force every building into a drone workflow.
It is to achieve the required inspection coverage by selecting an appropriate method for each site.
Conventional Drone → Power-Tethered Drone → Alternative Visual Inspection
The inspection requirement remains constant even when the access method changes.
Coordinating Multiple Field Teams
Scale also requires parallel operations.
The HDB technical specifications require at least six field teams, comprising pilots, safety personnel and ground staff or traffic controllers, supported by adequate drone equipment.
That changes project management significantly.
Flight schedules, manpower, equipment availability, weather, permit windows and reporting progress all need to be coordinated simultaneously.
NovaPeak therefore treats deployment planning as part of the inspection system itself.
Each completed flight is only one stage in a larger workflow that continues through data processing, engineering review and report acceptance.
From Drone Imagery to Structured Inspection Data
A large inspection programme can generate an enormous photographic dataset.
Without structure, thousands of images quickly become difficult to manage.
The programme therefore requires an integrated building façade inspection system in which drones act as the image and video capture platform while software is used to organise the collected data and assist in identifying façade types and defects.
The technical requirements specifically call for AI and machine-learning capability to support productivity, accuracy and repeatability in façade assessment.
Images are indexed for traceability using information such as location, geolocation, date and time of survey. The software workflow supports image organisation, automated defect detection, defect highlighting and review by the Competent Person.
This converts aerial photography into a structured inspection dataset.
AI-Assisted Defect Identification
AI plays a supporting role in the inspection process.
Rather than replacing engineering review, AI is used to help process large volumes of imagery and identify potential façade anomalies for further assessment.
The programme requires the inspection system to support autonomous defect detection using image-recognition and machine-learning methods, while still allowing the Competent Person to review, mark and annotate defects or anomalies before reporting.
This human-and-machine workflow is particularly relevant at scale.
AI helps reduce the burden of manually reviewing every image, while qualified professionals remain responsible for interpreting the results.
Capture → AI Screening → Human Review → Engineering Assessment
That distinction is central to NovaPeak’s approach to AI-enabled inspection.
Defect Traceability and Escalation
Defect identification is useful only when an observation can be traced back to the correct building and façade location.
The inspection workflow therefore associates images and identified defects with the relevant asset and location so they can be reviewed and incorporated into the final inspection record.
Potentially serious defects are also subject to escalation.
The project requirements specify that major defects that may affect public safety — such as dangling delaminated finishes or specified structural cracking — must be reported to HDB on the same day they are detected.
This means the programme is not merely an archive-building exercise.
Inspection information must be capable of triggering timely action when conditions warrant it.
Protecting Resident Privacy
Large-scale residential drone inspection introduces another important responsibility: privacy.
The programme requires images or video associated with residents or the interiors of residential premises to be masked, while unrelated information is removed following the flight.
Privacy protection therefore forms part of the data workflow rather than being treated as an afterthought.
For NovaPeak, responsible drone inspection requires both technical capability and data governance.
Engineering Review and Reporting
The final value of the programme does not come from the drone images themselves.
Each inspected block requires a professional report documenting the current façade condition, supported by relevant photographs, findings and recommendations.
The reports are reviewed and endorsed by an appointed Competent Person, who may be a suitably qualified Professional Engineer or Registered Architect with the required façade-inspection certification.
The reporting requirement also goes beyond showing defects alone.
HDB requires representative photographs of the general façade condition and expects the report to contain professional engineering views, assessment, judgement, conclusions and follow-up recommendations.
This reinforces an important principle:
the drone captures evidence; the professional determines its significance.
Managing a Multi-Year Inspection Programme
The programme spans two phases across 2026 and 2027.
Indicative quantities within the technical specifications comprise different combinations of conventional drone inspection, power-tethered drone inspection and routine visual inspection, with the actual quantities subject to adjustment.
Managing that volume requires continuous operational visibility.
Biweekly progress reporting tracks completed flights, inspected blocks, report-submission status and planned activities for the following two weeks. Deviations from the programme must be identified rather than allowed to accumulate unnoticed.
The result is effectively a production system for building inspection:
Quality Assurance at Portfolio Scale
Consistency becomes increasingly important as inspection volume grows.
The programme therefore includes requirements covering inspection-system documentation, software control, licensed pilots, flight and training records, workplace safety, operational risk assessment and ground-safety procedures.
For NovaPeak, this is one of the most important lessons from large-scale inspection.
Scaling inspection cannot mean relaxing inspection discipline.
In fact, the larger the programme becomes, the more important standardised workflows, repeatable evidence requirements and quality-control procedures become.
What the Programme Demonstrates
The HDB Façade Audit Inspection programme demonstrates NovaPeak’s ability to operate beyond individual building projects and deliver technology-enabled inspection at portfolio scale.
It brings together:
Drone Operations + Tethered Systems + AI-Assisted Analysis + Engineering Review + Data Governance + Regulatory Compliance + Programme Management
The value lies not in any single component.
It lies in connecting all of them into a repeatable inspection workflow.
From Drone Operations to Asset Intelligence
A programme involving more than 1,400 buildings also creates something that individual inspections cannot: a substantial structured dataset of real building conditions.
The project requirements include progressively annotated defect imagery that can support future model training, creating the potential for inspection systems to improve as additional validated data becomes available.
Over time, these structured datasets can support increasingly sophisticated capabilities in defect classification, condition comparison, predictive maintenance and portfolio-level asset intelligence.
The longer-term opportunity therefore extends beyond performing inspections faster.
It is about creating a digital evidence base from which better building-maintenance decisions can be made.