A bridge monitoring system consists of sensors which create a data pipeline, and of an analysis layer that translates the raw data coming from sensors into insights for civil engineers. This guide walks through the sensors, the methods, the architecture trade-offs, and the regulatory drivers behind modern bridge monitoring systems.

What a bridge monitoring system actually is

A system that monitors a bridge consists of three elements:

• a sensor network on the structure,
• a data pipeline that moves measurements off the bridge and into storage,
• and an analysis layer that turns raw time series into a condition judgement.

Such a system complements periodic inspections by offering data an inspection cannot provide, like slow modal drift over months, and transient overloads that happen between inspections and leave no visible mark. On the Chetwynd Bridge the system caught illegal overweight vehicles, something no quarterly walk-over would have seen.

What a monitoring system measures depends in large part on the failure mode that is most worrisome. Deck deflection and modal frequency track global stiffness loss, tilt tracks foundation movement and pier settlement, local strain tracks fatigue at a known hot spot, and vibration velocity matters when nearby construction or traffic is the concern rather than the bridge itself.

Monitoring and inspection are not substitutes

As above, continuous monitoring complements visual inspection and it cannot replace it. Sensors catch what inspectors miss, like slow degradation, or what happens between visits, the slow frequency drop and the overload event.

Sensors cannot detect hidden corrosion under paint, section loss, fresh impact damage or defects of human judgment. Therefore, both are required to really secure an infrastructural project.

Sensors involved in bridge monitoring

A bridge needs to be monitored by a small set of complementary instruments, each tied to a specific measure and a specific failure mode.

Accelerometers

A triaxial MEMS accelerometer measures acceleration in three axes, and from synchronised acceleration across the deck you recover the modal parameters, natural frequencies, mode shapes, and damping. Since these are global indicators of stiffness, the accelerometer is the basis for complete bridge monitoring. When stiffness drops because of cracking, corrosion or fatigue, the natural frequencies fall with it, which is why a long accelerometer record is the single most informative dataset on most bridges.

Move's DECKAXE-SHM samples from 40 to 640 Hz and feeds the MyMove Modal Analysis Tool directly. Sampling rate is an important factor, because in case it is too low, you alias the higher modes. Otherwise, when it is too high, you burn battery and bandwidth for frequencies a bridge will never really reach. For a medium-span road bridge the first few bending modes usually sit below 10 Hz.

Tiltmeters

A tiltmeter measures rotation, and rotation is the earliest visible symptom of a pier settling or an abutment moving. Sub-milliradian resolution is the threshold worth caring about, because anything coarser drowns the slow geotechnical signal in thermal noise. Move's tiltmeter hits sub-mrad and is the instrument we reach for first on any bridge where foundation movement is plausible.

Chained across several points, tiltmeters reconstruct a deflection profile of the deck during a load test without a single displacement transducer hanging over the river. That is the trick behind static deflection monitoring, and it is far cheaper to install than the alternatives.

Displacement sensors

When inferred deflection under dynamic load is not enough, a displacement sensor can be placed to measure that deflection directly and at high sampling rates.

On Chetwynd, it was the displacement sensors plus video that pinned down the overweight-vehicle events. Direct displacement is expensive in installation effort, so we use it where the deflection number itself is the deliverable, not as a default channel everywhere.

Vibrometers

A vibrometer measures peak particle velocity, which is the metric the construction-vibration standards are written around, and it is used to understand if construction work is shaking the structure past a safe limit.

Move's DECKVBR-STD runs 1 to 100 Hz and is compliant with DIN 4150-3, UNI 9916 and BS 7385. On a bridge it serves when there is piling, demolition or heavy plant working nearby.

Strain gauges and fibre optic, through a node

Strain and fibre-optic sensing are local methods, and damage is usually local, so they are used at known fatigue hot spots.

They are not Move's primary line, but our communication node digitises third-party analog sensors and brings them into the same wireless network and the same platform, so analysis for the whole project can be done in the same tool.

Environmental sensors

Temperature is the great noise generator in monitoring, since a natural frequency can shift several percent across a day purely from thermal effects on stiffness and bearing behaviour, which is enough to look like damage.

By co-locating temperature and environmental measurement, the environmental data can be used to separate the daily thermal fluctuations from a downward trend.

Turning data into a diagnosis

There are three types of analysis used for bridge monitoring that use the data coming from a well-structured sensor system.

Vibration-based monitoring and operational modal analysis

Operational modal analysis, or OMA, is the method that fits in-service bridges. It extracts modal parameters from the structure's response to the excitation it already gets from traffic, wind and microtremor. This method doesn't require a shaker, road closures or controlled input.

Experimental modal analysis with a forced input is more precise, but on a live bridge the cost and traffic disruption of an eccentric-mass shaker make it impractical except for a one-off commissioning test.

The peer-reviewed OMA literature is now extensive and the algorithms, frequency-domain decomposition and stochastic subspace identification among them, are settled.

Static deflection and load testing

Static methods help define the total capacity. A static load test puts known loads on the deck and measures the deflection, which then validates the structural model and the load rating.

Tiltmeter chains and displacement sensors are used for these load tests, and the MyMove Static Deflection Tool can be used to later reconstruct the profile.

Dynamic and static testing are complementary, since the dynamic record tells you how the bridge is ageing, while the static test tells you what it can carry.

Wired vs wireless, and how many sensors

A wireless system is almost always the better choice for retrofits, which means installing new monitoring systems on existing infrastructure. That's because a wireless system requires only a few devices to be installed in specific points of the structure, which usually means just a few bolts.

Cabling an existing bridge for a wired system means trenching, conduit, weatherproofing and a cost per linear foot that makes the cable become more expensive than the sensors it serves. The published comparisons put wireless installation at well under half the cost of the wired equivalent.

Sparse and well-placed beats dense and badly-placed

A typical sensible first deployment on a medium-span bridge is four to eight triaxial accelerometers at deck mid-span and the quarter-spans, plus one tiltmeter per pier, and an environmental channel for temperature correction. That array recovers the first several modes, catches foundation movement, and produces a dataset still interpretable for a single engineer. A system that uses too many sensors can actually cause a lot of noise in the data and create unnecessary difficulties. In general, sensors should be added when the data is showing points that are underrepresented.

What the regulations actually require

Regulation is now one of the main reasons bridge monitoring systems get funded. Overall, governments are realising that most of the current infrastructure in Europe and North America is ageing, and without a strategic response to this problem, tragedies like the collapse of the Morandi Bridge in Genova, Italy in 2018 could start repeating too often.

In Italy the MIT 2020 Guidelines on classification and risk management of existing bridges set up a multi-level assessment under which the higher-risk structures effectively need instrumented monitoring.

In the United States the framework is the FHWA's National Bridge Inspection Standards, which mandate inspection intervals rather than continuous monitoring, but which increasingly recognise instrumented data as a way to extend intervals and target special inspections.

On the vibration side, DIN 4150-3, UNI 9916 and BS 7385 set the peak-particle-velocity thresholds that govern when nearby work is shaking a structure too hard. These are well-written standards for what they cover, but note the scope. They address vibration impact on the structure, not the structure's own modal health, so a vibrometer compliant with DIN 4150-3 answers the construction-impact question and tells you nothing about long-term stiffness loss.

How we approach bridge monitoring at Move

At Move Solutions, we deploy wireless sensors on site, which connect over LoRaWAN and push data to the MyMove cloud, where the analysis tools can be used in one single platform that creates reports and stores them for compliance.

The Chetwynd Bridge shows what the constraints look like on a protected structure. No bolts and no adhesives were allowed on the listed ironwork, so the sensors were mounted with high-strength magnets. There was no room for a solar array, so the team charged a backup power unit from the existing 24V street lighting. The instruments were even painted to match the bridge so they would clear heritage approval. Six months of accelerometer data, analysed in February 2024, established the modal baseline, and the record has since been used to justify the case for a replacement bridge by documenting gradual deterioration that left no visible mark.

The Zambeccari Bridge in Pontremoli, an early-1900s reinforced-concrete structure reopened with a reduced carriageway after a closure, runs continuous remote monitoring set up with Vega Engineering. The Bridge of the Gods in Oregon, deployed with Parsons, brought the same wireless approach to a historic Columbia River crossing, capturing stress, vibration and movement on a cantilever truss.

Frequently Asked Questions

Is wired monitoring ever the right choice over wireless?

Yes, on new long-span bridges instrumented during construction, where cable trays are already going in and you need perfect synchronisation and continuous power across many high-rate channels. Even then though, we see that sometimes clients need a mixed system of wired and wireless sensors and they use a platform like MyMove for analysis.

How does temperature affect modal frequency readings, and does it cause false alarms?

Strongly. Thermal effects on stiffness and bearings can move a natural frequency by several percent over a single day, which is the same order as a real damage signal. Without a co-located temperature channel and a correction model, an alarm threshold will either fire every warm afternoon or sit so wide it misses genuine drift.

Can a bridge monitoring system replace scheduled inspections?

No, continuous monitoring complements inspections. Whereas sensors catch slow modal drift and transient overloads between visits, inspectors catch hidden corrosion, section loss and impact damage that sensors are not equipped to measure. The regulatory direction, including the Italian MIT 2020 guidelines, treats monitoring and inspection as a combined regime rather than one substituting for the other.

‍