TL;DR
Structural monitoring prevents failures by measuring movement before it becomes damage — tracking deflection, settlement, tilt and crack growth against pre-defined trigger levels so engineers can intervene while the structure is still safe. On Australian industrial sites this means survey-grade total stations, automated monitoring stations and prisms reading to sub-millimetre precision, GNSS on tailings dams, and laser scanning to map deformation across an entire structure. The aim is simple: turn an invisible, accelerating movement into a number on a dashboard that triggers action days or weeks before a wharf, headframe or embankment reaches the point of collapse.
Key takeaways
- Structures rarely fail without warning — they accelerate. Monitoring captures the velocity and acceleration of movement (mm/day, then mm/hour) so a green-amber-red trigger system can drive evacuation or load reduction before the failure point.
- Automated total stations (Leica TM60/Nova MS60, Trimble S9 HP) reading networks of monitoring prisms achieve repeatability of roughly ±0.3–1 mm at 100–200 m, captured on cycles from hourly down to a few minutes.
- Tailings storage facilities are now monitored against the Global Industry Standard on Tailings Management, typically combining survey prisms, GNSS, InSAR and instrumentation — failures like the 2019 Brumadinho dam show how fast a TSF can move from stable to catastrophic.
- Trigger Action Response Plans (TARPs) only work when the deformation tolerance is set by a structural engineer against the real structure, expressed in GDA2020/MGA2020 and AHD, not borrowed from a generic spec.
- A monitoring programme is a fraction of the cost of a single failure — automated systems run from roughly $1,500–$5,000/month against unplanned shutdowns and remediation that routinely reach seven figures.
Why structures fail — and why it is preventable
Industrial structures almost never fail at random. A wharf does not collapse the moment a ship berths heavy; a tailings embankment does not breach on a still day with no warning movement; a mine headframe does not buckle without first leaning. In nearly every case the structure has been moving — slowly, then less slowly — for days, weeks or months before the event. The failure is not the movement; the failure is that nobody was measuring the movement, or nobody had agreed what number meant "stop".
That is the entire premise of structural monitoring. It converts a hidden, progressive process into a measured one. A column that settles 2 mm a year is doing nothing alarming. The same column settling 2 mm in a week is telling you the ground beneath it is failing. Without monitoring, both look identical — the structure simply appears "fine" until the day it is not.
The economics make the case on their own. The mining and resources sector employs close to 300,000 people directly and exports around AUD $385 billion a year; a single unplanned shutdown of a major processing line, conveyor or export wharf is measured in millions of dollars per day of lost throughput. Against that, continuous structural monitoring is one of the cheapest insurance policies a site can buy. The question is rarely whether monitoring is worth it — it is whether the monitoring is designed to actually catch the failure mode that matters.
What structural monitoring actually measures
"Structural monitoring" is not one measurement — it is a set of measurements chosen to suit the structure and its likely failure mode. The core quantities are:
- Deflection and displacement — movement of a point in 3D (easting, northing, height) relative to a stable reference, reported in GDA2020/MGA2020 horizontally and AHD vertically.
- Settlement and heave — vertical movement, typically the controlling parameter for foundations, tanks, silos, wharves and embankments.
- Tilt and rotation — angular change, critical for headframes, chimneys, retaining walls and high-mast structures, often captured by tiltmeters as well as survey.
- Convergence — the closing-in of an opening, the key measurement in underground mining drives, tunnels and shafts.
- Crack movement — opening, closing and shear across an existing crack, measured with crackmeters or repeat survey.
- Strain and load — measured by strain gauges and load cells where the structural engineer needs the force, not just the geometry.
The skill is not in capturing one of these — it is in capturing the right combination at the right frequency, and crucially in capturing the rate of change. A single displacement reading tells you where a point is. A time series tells you where it is going and how fast — and rate of change is what every meaningful trigger level is built on.
The monitoring technologies and what each is for
| Technique | Typical precision | Best suited to | Representative equipment |
|---|---|---|---|
| Automated total station (ATS) + prisms | ±0.3–1 mm at 100–200 m | Wharves, headframes, plant structures, bridges, pit walls | Leica Nova TM60/MS60, Trimble S9 HP |
| Manual precise total station | ±1 mm + 1 ppm | Periodic deformation surveys, shutdown checks | Leica TS16, Trimble S7 |
| GNSS continuous | 3–8 mm (real-time), ~2–5 mm post-processed | Tailings dams, large slow-moving slopes, remote assets | Leica GMX910, Trimble monitoring receivers |
| Terrestrial laser scanning | 2–4 mm at 50 m | Whole-of-structure deformation, ovality, as-built vs design | Leica RTC360/P50, FARO Focus, Trimble X12 |
| UAV/drone photogrammetry & LiDAR | 15–50 mm | Tailings surfaces, stockpiles, inaccessible faces, broad-area change | DJI Matrice 350 RTK + L2, Trimble UX5 |
| Geotechnical instrumentation | sensor-dependent | Pore pressure, tilt, strain, crack width | Tiltmeters, piezometers, crackmeters, load cells |
No single technology is "best". An automated total station gives you sub-millimetre movement on hundreds of discrete prisms but only sees what you put a prism on. A laser scan gives you millions of points across an entire structure but at lower precision and lower frequency. GNSS works in the open with no line of sight but struggles at sub-centimetre real-time. UAVs cover ground no other method can reach but at coarser precision. A well-designed programme layers them: ATS or GNSS for high-frequency point movement and triggers, periodic scanning for whole-of-structure context, drones for broad-area surface change.
CASA Part 101 governs the drone component — beyond visual line of sight, controlled airspace near ports and aerodromes, and operations over people all carry specific approvals, and any reputable provider flies with a Remote Operator's Certificate and current Remote Pilot Licences.
How monitoring prevents failure: the trigger system
This is the mechanism that actually does the preventing. Measurement alone changes nothing — it is the Trigger Action Response Plan (TARP) that turns a number into an intervention.
A TARP defines, before monitoring starts, what movement is acceptable and what each threshold requires. A typical three-tier structure looks like this:
| Level | Example trigger | Response |
|---|---|---|
| Green (normal) | < 5 mm cumulative, < 0.5 mm/day | Routine review of data; no action |
| Amber (alert) | 5–15 mm cumulative, or 0.5–2 mm/day sustained | Increase reading frequency, notify structural engineer, inspect, plan intervention |
| Red (alarm) | > 15 mm or accelerating > 2 mm/day | Restrict or cease loading, evacuate exclusion zone, engineering intervention |
The numbers above are illustrative — and that is the single most important point. The trigger levels must be set by the structural or geotechnical engineer responsible for the asset, against the real structure's capacity, geometry and failure mode. A 20 mm deflection that is trivial for a long-span conveyor gantry may be catastrophic for a brittle masonry retaining wall. Borrowed or generic triggers are worse than none, because they create false confidence.
What makes the trigger system genuinely preventive is automation. A modern automated monitoring station reads its prism network on a fixed cycle — every few minutes for high-risk structures — applies atmospheric corrections, computes displacement against the reference network, and pushes the results to cloud software (Leica GeoMoS, Trimble 4D Control or equivalent). When a point breaches amber, the system emails and SMS-alerts the nominated engineers automatically, day or night. When it breaches red, it can trigger sirens, beacons and gate controls. The failure is prevented not by the surveyor watching a screen, but by the system escalating faster than a human could.
Where it matters most on Australian sites
Three categories dominate structural monitoring demand in Australian industry.
Tailings storage facilities (TSFs). Following Brumadinho (2019) and Mount Polley, dam monitoring is now governed by the Global Industry Standard on Tailings Management, and Australian operators have moved to layered, continuous monitoring — GNSS and survey prisms on the embankment crest and downstream face, piezometers for pore pressure, InSAR for broad-area surface velocity, and drones for routine surface mapping. A TSF can move from imperceptible creep to runaway failure in hours, so the value of high-frequency, automated movement data here is at its absolute highest.
Wharves, ship loaders and export infrastructure. At Port Hedland, Dampier, Gladstone, Newcastle and Hay Point, berths and ship loaders carry enormous cyclic and impact loads in a corrosive marine environment. Monitoring catches pile settlement, beam deflection and crane rail movement before a structural defect becomes an outage that backs up an entire export chain.
Underground and open-pit mining structures. Pit-wall and slope monitoring (radar plus survey prisms), shaft and decline convergence, and headframe tilt are all standard. In an active pit, the difference between a controlled batter failure and a fatality is measured in the hours of warning the monitoring system buys.
Plant structures — silos, bins, conveyor gantries, pipe racks, furnace and boiler steelwork — round out the picture, particularly where settlement, vibration fatigue or thermal cycling is suspected.
Setting up a monitoring programme that works
A programme that prevents failures rather than just collecting data follows a clear sequence.
- Define the failure mode with the engineer. What is the structure most likely to do, and what movement signals it? This drives every later decision.
- Establish a stable reference network. Movement is meaningless without a datum that is not itself moving. Reference monuments are placed well outside the influence zone, tied to GDA2020/MGA2020 and AHD, and themselves checked for stability.
- Capture a baseline. Several reading epochs before any "change" is declared, so natural daily and seasonal variation (thermal expansion in particular) is understood and not mistaken for structural movement.
- Choose technology and frequency to match the rate of the failure mode. A slow-moving foundation may need weekly survey; a high-risk pit wall or TSF needs continuous automated readings.
- Set and document the TARP. Green/amber/red triggers, named responders, and the specific action at each level, signed off by the responsible engineer.
- Automate alerting and review. Cloud dashboards, automatic alarms, and a scheduled human review of trends — because slow drift that never trips a daily-rate trigger still matters over months.
Watch out: the most common way a monitoring programme fails to prevent a failure is thermal noise. Steel structures move several millimetres a day purely from the sun. If your baseline and corrections do not account for temperature, you will either chase phantom alarms or, far worse, set your triggers so wide that real movement hides inside the daily thermal swing.
Cost considerations
| Cost factor | Typical range (AUD) | Notes |
|---|---|---|
| Periodic deformation survey | $2,500–$8,000 per epoch | Manual total station; suits slow movement |
| Automated monitoring station install | $25,000–$80,000+ | ATS, prisms, power, comms, software setup |
| Automated monitoring — ongoing | $1,500–$5,000 per month | Data hosting, alarm management, engineering review |
| Laser scan deformation survey | $4,000–$15,000 | Whole-of-structure, repeatable for change detection |
| GNSS array (per receiver) | $8,000–$20,000 installed | Tailings dams, remote slopes |
Set against a failure, these figures are small. A breached tailings dam, a collapsed wharf or a buckled headframe carries remediation, lost production, regulatory and — in the worst cases — human costs that dwarf any monitoring budget. The cost question is not "can we afford to monitor"; it is "what is the cheapest monitoring design that reliably catches the one failure mode that would actually hurt us".
Frequently asked questions
How accurate does structural monitoring need to be?
Accurate enough to detect meaningful movement well before the failure threshold. For most industrial structures that means sub-millimetre to low-millimetre repeatability, which automated total stations and precise levelling deliver. The target is not raw accuracy for its own sake — it is being able to distinguish real structural movement from noise (thermal, atmospheric, instrument) at the trigger levels the engineer has set.
How often should readings be taken?
It depends entirely on how fast the structure can fail. A stable foundation under observation may be surveyed monthly. A suspect retaining wall might be read daily. A high-risk pit wall or active tailings facility is monitored continuously, with automated cycles down to minutes. The rule is that your reading frequency must be much faster than the time it would take the structure to move from amber to failure.
Who sets the trigger levels?
The structural or geotechnical engineer responsible for the asset, against its actual capacity and failure mode — not the surveyor, and never copied from a generic template. The surveyor's job is to deliver movement data reliable enough that those engineer-defined triggers can be trusted to fire at the right time.
Can monitoring be done without disrupting operations?
Yes. Automated systems run continuously with the plant fully operational — prisms are small, fixed and unobtrusive, GNSS antennas sit out of the way, and laser scanning and drone work capture data from a distance. That is a core advantage: monitoring watches the structure under real working load, which is exactly when failures occur.
What standards apply in Australia?
Spatial data is referenced to GDA2020/MGA2020 horizontally and AHD vertically. Tailings facilities are monitored against the Global Industry Standard on Tailings Management. Drone monitoring operates under CASA Part 101. Specific deformation tolerances come from the relevant AS/ISO structural and geotechnical standards and the asset owner's engineering requirements.
Request a structural monitoring assessment
Failures are prevented in the weeks before they happen, not in the moment they occur — and only if someone is measuring the right movement against the right trigger. Industrial Spatial Solutions designs and operates structural monitoring programmes across Australian mining, port and heavy-industrial sites, from periodic deformation surveys through to fully automated, alarmed monitoring of tailings dams, wharves, headframes and plant structures using Leica and Trimble survey systems, FARO laser scanning and CASA-compliant UAV survey. If you have a structure you are not certain about — a wall that may be leaning, a wharf carrying more than it was built for, an embankment that needs watching — call us on 0407 057 015 to scope a monitoring programme built around your structure's real failure mode and to request a quote.