A construction site changes the physical world while people continue living beside it.
Piles enter the ground.
Excavations remove support.
Groundwater conditions may change.
Retaining walls move under load.
Roads, utilities and neighbouring buildings remain where they were before the project began.
Monitoring is how the project turns some of those changing conditions into evidence.
Construction monitoring measures selected aspects of ground, structure, water and vibration over time so project teams can compare real behaviour with expected behaviour and respond before an unacceptable condition becomes damage.
BCA’s current pre-construction survey guidance states that before demolition, piling, foundation, tunnelling or site-formation works, builders must conduct surveys to establish the condition of nearby existing buildings and structures. BCA’s permit guidance also includes records for supervision of piling and pile-load tests together with monitoring of building settlement.
BCA’s current ERSS resources further include a performance-based impact-assessment framework for excavation and tunnelling works and an Observational Method framework that connects monitoring to pre-planned engineering responses.
This article explains the monitoring logic. It does not provide alarm values, instrument spacing, settlement limits, vibration limits or emergency actions. Those are project-specific and belong to the approved engineering and regulatory framework.
For the activities being monitored, read How Excavation and Earth-Retaining Systems Work on HDB Construction Sites and How Groundwater Control Works During HDB Excavation. For the full public-housing system, return to How HDB Works in Singapore.
The first measurement is often taken before construction changes anything
If a crack is found in a neighbouring building after excavation begins, one immediate question is whether the crack existed before the work.
That is why pre-construction surveys matter.
A pre-construction survey establishes a documented before-state of nearby buildings and structures. Depending on the scope, it can record visible conditions such as cracks, finishes, deformation or other relevant observations before major work begins.
The survey does not predict the future. It protects the evidence boundary between pre-existing condition and later change.
Without that baseline, later arguments can become memory against memory.
Condition and movement are different forms of evidence
A photograph shows condition at one moment.
A settlement marker or survey prism can show how a point changes over time.
Both are valuable, but they answer different questions.
- Condition record: What was visible before or at a particular time?
- Movement record: How did a measured point change relative to a reference?
A building can move slightly without developing visible damage. Visible damage can also have causes unrelated to the new construction.
Monitoring becomes powerful when several kinds of evidence can be compared instead of asking one photograph to explain everything.
Settlement monitoring asks whether a point is moving vertically
Settlement markers or survey points can be used to track vertical movement at selected locations on buildings, roads, structures or ground.
The measurement requires a stable reference system. A point can only be described as moving if the reference used to measure it remains trustworthy.
This is why construction monitoring depends on construction surveying. The same geometry that places the new building also helps measure change around it.
One settlement reading says where a point is now. A series of readings says whether a trend exists.
Trend often matters more than one isolated reading
Imagine a hypothetical monitoring point whose readings vary slightly around the same level because of measurement uncertainty.
Now imagine another point whose readings move steadily in one direction over several excavation stages.
The second dataset may deserve more attention even if no single reading looks dramatic.
This is why construction monitoring is a time-series problem.
Competent interpretation considers magnitude, rate, direction, construction stage, instrument reliability and the design’s expected behaviour.
Retaining-wall movement can be measured below the surface
Deep excavation walls can deform under soil and groundwater pressures.
Instrumentation such as inclinometers can be used in suitable projects to measure lateral movement with depth.
This is useful because the maximum wall movement may not occur at the ground surface.
A surface survey point can therefore tell one part of the story while a subsurface instrument tells another.
The instrument type, depth, interpretation and trigger criteria are specialist project decisions.
Groundwater instruments measure another hidden state
Piezometers and other groundwater instruments can measure selected groundwater levels or pressures.
These readings matter because excessive drawdown can contribute to ground settlement outside an excavation, while rising pressure can create other excavation risks.
Read How Groundwater Control Works During HDB Excavation for the hydraulic mechanism.
Monitoring becomes especially useful when groundwater readings and movement readings are considered together.
Vibration monitoring owns a different question from settlement monitoring
Piling, demolition or other construction activities can generate vibration.
A vibration monitor measures aspects of dynamic motion over time. A settlement survey measures longer-term positional change.
The two should not be confused because a building can experience vibration without permanent settlement, and settlement can occur through processes that do not produce obvious vibration.
The correct project-specific limits and measurement methods depend on the relevant standards, receptors and construction activities.
A monitoring instrument is useful only if its own position is trustworthy
An instrument can be damaged, disturbed or read incorrectly.
A settlement marker that is knocked loose can appear to show building movement when the real change is at the marker itself.
A piezometer can produce misleading data if installation or interpretation is wrong.
Monitoring therefore needs instrument-quality control as well as data collection.
The same rule applies throughout engineering: a surprising measurement should be investigated before it is either dismissed or treated as unquestionable truth.
Frequency matters because construction changes by stage
One reading before excavation and one after completion can miss what happened in between.
Monitoring frequency may need to change with construction activity, risk and observed behaviour.
A deep excavation stage, major dewatering change or critical support transition may justify closer observation than a quiet period under stable conditions.
The actual frequency belongs to the project’s monitoring plan.
The important principle is responsiveness: collect data often enough that the project can act before the next construction decision outruns the evidence.
Monitoring needs trigger levels before the readings arrive
If a team waits until an instrument moves and then asks what amount of movement is concerning, the monitoring system is incomplete.
Well-designed monitoring frameworks define expected response, trigger levels, reporting routes and actions in advance.
The exact numbers are project-specific and are deliberately not reproduced here.
Conceptually, projects can use escalating states such as normal, alert and action conditions, with different responses at each level.
The value of a trigger is that the project does not have to invent its risk appetite under schedule pressure.
A trigger is not a prophecy of failure
Crossing an alert level does not necessarily mean a building is unsafe.
It means the observed behaviour has reached a condition that the monitoring plan says deserves a defined response.
That response can include:
- checking the instrument and reading;
- increasing monitoring frequency;
- reviewing nearby construction activity;
- checking groundwater or other related data;
- engineering assessment;
- altering or pausing work where the approved plan requires it.
The list is conceptual, not a universal response protocol.
Monitoring works best when several instruments tell a coherent story
One instrument can be wrong.
One point can also experience a local condition that is not representative of the wider site.
Confidence improves when independent observations support one another.
For example, a wall inclinometer may show increasing lateral movement while nearby settlement points and groundwater readings also change in a physically consistent way.
Alternatively, one isolated reading may jump while every related measurement remains stable, suggesting the first job is to verify the instrument or local condition.
Engineering interpretation searches for coherence without forcing all data to agree artificially.
The Observational Method formalises this evidence loop
BCA’s Observational Method framework allows measured performance to inform construction decisions under a pre-planned design framework.
The method is valuable because geotechnical systems contain uncertainty. Rather than pretending every parameter is known perfectly, the project can define conservative expectations, monitor real behaviour and adopt predefined alternatives where the observed response permits or requires them.
The method depends on discipline:
- monitoring must be reliable;
- decision limits must exist in advance;
- alternative actions must be engineered before they are needed;
- responsibilities must be clear;
- data must reach decision-makers in time.
Without those controls, “observational method” would collapse into improvisation.
The fastest safe response depends on fast information flow
Monitoring can be technically excellent and operationally weak if data take too long to reach the responsible engineer.
A reading may be collected.
A report may be prepared later.
Construction may meanwhile continue through another stage.
Digital monitoring systems can reduce that delay by transmitting data, plotting trends and issuing alerts. The benefit is not the dashboard. It is the shorter time between a meaningful change and a competent response.
This is the same logic that makes Integrated Digital Delivery valuable elsewhere in construction.
Automatic alerts should not automate engineering judgement
A monitoring system can compare a reading with a threshold and send an alert automatically.
The alert still needs interpretation.
Was the sensor disturbed?
Did construction enter a new stage?
Do related instruments show the same pattern?
Does the approved response require a pause, an inspection, more readings or another action?
Automation accelerates attention. It does not own the professional decision.
Monitoring nearby buildings protects both safety and evidence
Where nearby buildings could be affected by excavation or foundation work, monitoring has two important public roles.
- Safety: detect movement early enough for the project to respond.
- Evidence: preserve a measured record of what actually changed during the works.
Those roles support both residents and builders.
A neighbour should not have to rely solely on a project’s reassurance that “nothing moved”.
A builder should not have to accept responsibility for every old crack simply because construction is now nearby.
Pre-construction condition records and monitoring provide a stronger factual basis for investigation.
Impact assessment connects predicted movement to neighbouring vulnerability
Not every building responds the same way to the same ground movement.
A rigid modern structure, an older masonry building, a road and a buried utility have different sensitivities.
BCA’s impact-assessment framework reflects this by requiring projects to consider adjacent buildings when assessing ERSS and tunnelling effects.
The monitoring strategy should therefore be connected to what could be harmed, not merely to what is easiest to instrument.
A number becomes meaningful when it is related to the vulnerability of the asset it is intended to protect.
Monitoring itself can influence construction sequence
If the project requires a stable reading or engineering review before entering the next excavation stage, monitoring becomes part of the critical path.
This can look like delay from the outside.
The important distinction is whether the pause protects a real decision gate or merely reflects slow information handling.
Good project management reduces the second while preserving the first.
Wintour v1.0 applies the same release logic editorially: evidence gates do not disappear because throughput is inconvenient.
Instrumentation should survive the construction around it
Monitoring devices are installed in an environment full of excavators, deliveries, vibration, weather and changing access.
The project therefore needs to protect instruments from accidental damage and preserve access for reading or maintenance.
A damaged instrument can create two losses at once:
- the device itself is lost;
- the continuity of the historical record may be broken.
Replacing the device does not automatically recreate the missing data.
A long stable trend can be valuable evidence too
Monitoring is often discussed only when something moves.
Stable readings can also be useful because they show that selected monitored quantities remained within expected behaviour through particular construction stages.
This can support safe continuation and refine confidence in the design model.
The absence of movement at one point does not prove nothing moved anywhere. Again, scope matters.
Monitoring can support optimisation without lowering safety
BCA’s Observational Method framework is explicitly intended to support construction efficiency where monitored performance supports an optimised design response.
This is a good example of evidence allowing flexibility.
The project does not lower the safety objective.
It uses measured behaviour to decide whether a conservative assumption can be refined within a pre-approved framework.
The distinction matters. Optimisation based on evidence is engineering. Optimisation based on hope is schedule pressure.
Residents should not be asked to interpret project instruments themselves
A marker, prism or instrument visible near a building can create concern.
The public does not need to infer the meaning from appearance.
A useful project communication can explain:
- what general type of monitoring is being undertaken;
- which activity it relates to;
- that readings are reviewed by the responsible project team;
- where residents can report observations or concerns.
That communication should not expose sensitive operational thresholds casually or transfer monitoring responsibility to residents.
Monitoring is the project’s control system. Public understanding should make the system more transparent, not outsource it.
A resident’s observation can still be valuable evidence
Instrumentation does not make human observation irrelevant.
A resident may notice a new crack, sticking door, unusual vibration or another change not captured directly by the project’s sensor locations.
A useful report records:
- location;
- date and time;
- what changed;
- whether it appears to be growing or recurring;
- photographs where safe and appropriate.
The resident should not enter the construction site or attempt structural diagnosis.
The observation becomes another input for the responsible parties to compare with baseline records and monitoring data.
Failure mode: collecting data nobody reviews promptly
A monitoring system that records yesterday’s problem after today’s excavation stage has already advanced is operationally weak.
Failure mode: one instrument becoming the whole truth
Instrument error, local conditions and limited spatial coverage mean important decisions should consider the appropriate wider evidence.
Failure mode: changing trigger levels after an uncomfortable reading
Trigger frameworks should be governed through the approved engineering process, not moved casually to protect programme convenience.
Failure mode: treating a trigger exceedance as proof of structural danger
A trigger is a decision threshold within a monitoring plan. The appropriate engineering response determines what the observation means.
Failure mode: recording the before-state poorly
Weak pre-construction records reduce the project’s ability to distinguish pre-existing conditions from construction-related change later.
Failure mode: assuming monitoring prevents movement
Monitoring observes. It does not physically support the ground.
ERSS, groundwater controls, construction sequence and other engineering measures control the physical system. Monitoring tells the project how that control is performing.
A better construction-monitoring reading
- What was the nearby condition before work started?
- Which construction activity could create movement or vibration?
- What quantity is each instrument actually measuring?
- Where is the instrument located relative to the risk?
- What is the expected trend?
- How frequently is the data reviewed during critical stages?
- What trigger framework governs the response?
- What related instruments or observations confirm or challenge the reading?
- Who has authority to change or stop the work if required?
- How is the record preserved for later verification?
Follow one hypothetical excavation through the monitoring loop
Before excavation starts, nearby buildings are surveyed and monitoring points are established.
The first excavation stage begins.
Wall-movement instruments, settlement points and groundwater readings remain within the expected range.
The project proceeds to the next approved stage.
Later, one settlement trend increases faster than expected.
The monitoring system flags the condition.
The team verifies the instrument, reviews adjacent points and groundwater data, compares the change with recent excavation and pumping activity, and carries out the engineering response required by the approved plan.
Work does not continue merely because the next contractor is waiting.
When the condition has been understood and the required controls are restored or confirmed, the project can proceed under the responsible engineering decision.
The monitoring system has not proved that nothing can go wrong.
It has shortened the distance between physical change and informed action.
The deeper construction principle
A deep excavation or foundation system is designed using a model of the ground and expected behaviour.
Monitoring creates a return path from the real site back into that model.
The project predicts.
The ground responds.
The instruments observe.
Engineers interpret.
The construction method adapts where the approved framework requires it.
Monitoring is valuable because it makes a hidden physical response visible early enough to change the next decision.
The deepest answer
Construction monitoring protects buildings around HDB sites by creating evidence across time.
Pre-construction surveys establish the before-state.
Settlement points measure vertical movement.
Inclinometers can measure lateral movement within retaining systems.
Piezometers observe selected groundwater conditions.
Vibration monitors own dynamic disturbance.
Trigger frameworks connect readings to action.
Impact assessment connects movement to the vulnerability of nearby assets.
The instruments do not make the excavation safe.
They make its behaviour observable enough that the people responsible for safety can act before an unexpected trend becomes an unacceptable consequence.
Sources and continuing through the HDB construction sequence
Singapore sources checked on 5 September 2026 include BCA’s pre-construction survey guidance, Permit for Structural Works records, performance-based impact-assessment framework and Observational Method framework.
Return to How Excavation and Earth-Retaining Systems Work on HDB Construction Sites, How Groundwater Control Works During HDB Excavation, or the complete How HDB Works in Singapore guide.