Wireless monitoring is replacing manual structural inspection in the same way that continuous patient telemetry replaced the twice-a-day ward round. The round still needs to happen, but now doctors have a better understanding of what happens to the patients throughout the day. For bridges, retaining walls and heritage buildings, the inspection interval, the gap between visits, is where most of the interesting structural behaviour takes place.

Manual inspection is not perfect

Under the US National Bridge Inspection Standards, the baseline routine interval is two years, extendable to four under an approved risk-based program. Italy's MIT 2020 guidelines run a tiered visual census on a similar cadence. Two years is a long time for a fatigue crack to propagate or for a pier to settle. If an inspector notices a crack after two years, there is no way to find out how long it's been there or how fast it's growing.

The FHWA's own reliability study on visual inspection of highway bridges found that only about two thirds of condition ratings land within one point of the reference rating, and that local defects such as cracks are routinely missed.

Reaching the structurally important points on a bridge is also expensive. A motorway lane closure in the UK runs upward of GBP 1,500 a day in traffic management for any engineer to touch the structure, and underwater pier inspections cost more again. Because of this, owners are incentivised to push toward the longest interval the regulator will allow, which widens the very gap that makes the snapshot unreliable.

Visual inspection is absolutely needed and it is good that it is being required by new regulation. But this method has blind spots that need to be filled.

Eliminating the blind spots of manual inspection

By installing a permanent network of sensors, property managers can turn a single data point into a continuous record that then helps not only to monitor but also to create a solid documentation of maintenance. With constant monitoring, a structure is measured under the loads it carries daily, through thermal cycles and traffic peaks and the occasional overweight vehicle that nobody logged.

From snapshots to continuous data

A DECKAXE-SHM accelerometer samples deck response at up to 640 Hz when it needs to and tracks the structure's modal frequencies over the years. Those frequencies serve as a global stiffness indicator. When this stiffness drops because of cracking, corrosion or fatigue, the natural frequencies drift downward, and a network watching continuously sees that drift before cracks start showing.

Slow phenomena are handled by a different sensor on the same network. A wireless tiltmeter at sub-milliradian resolution will register a pier rotating by fractions of a degree across a wet winter, which is a settlement signal impossible to see for an inspector from the deck.

Why wireless, and not wired

Continuous data does not require cables anymore since the newest sensor technology offers wireless sensors with batteries that last years. COST Action TU1406 found that about half of Europe's bridge stock predates 1980 and was never designed with sensor conduit in mind. This means most of these bridges would need to be retrofitted in order to connect wired sensors, and that would be the biggest cost of installing the system. Running cable to a pier base sitting in a watercourse, or along a deck soffit thirty metres up, means scaffolding, lane closures and waterproofed penetrations at every junction.

Wireless MEMS sensors on a LoRaWAN or mesh backbone remove that constraint. The sensor needs a stable mounting surface and a clear radio path to a gateway that can sit hundreds of metres away, nothing more. On the Scrivia river railway bridge, a seven-span structure around 160 metres long, we covered the full length from a single gateway with no inter-span cabling. On the Vespucci Bridge in Florence we placed sensors at pier and deck positions that would have demanded disproportionate cable infrastructure to reach. Sensor placement follows structural logic instead of cable routing, which is the whole point.

What sensors catch that inspectors never will

The first element that only sensors can catch is slow modal drift. A 0.5% shift in a fundamental frequency over eighteen months is invisible to the eye and unmistakable to the Modal Analysis Tool in MyMove, which tracks frequencies, mode shapes and damping across the deployment. On Chetwynd Bridge in the UK we ran exactly this kind of continuous modal monitoring, watching the structure's dynamic signature rather than waiting for a defect to surface.

The second is the transient overload. An overweight truck, a vehicle impact on a parapet, a seismic micro-event. These happen in seconds, almost always when no inspector is present, and they leave a peak in the acceleration or vibration record that explains a later anomaly. The Zambeccari Bridge deployment in Italy used wireless accelerometers for modal analysis on exactly this principle, capturing the structure's behaviour under real traffic rather than under a one-off commissioning test.

The third is environmental loading you would otherwise never quantify. In Odesa, a volunteer engineering team led by Emmanuel Durand has used our wireless sensors and MyMove to monitor vulnerable heritage monuments through wartime, capturing blast-related vibrations and tilt trends that an inspection schedule could never have been timed to catch.

What inspectors catch that sensors never will

A lot of phenomena cannot be monitored by sensors, such as hidden corrosion of reinforcement, chloride ingress or spalling concealed behind a soffit. These are things a trained inspector sees because they do not show up in data. A wireless network might flag that something changed and point the inspector at the right span, but the diagnosis of what and why still needs eyes, a hammer and judgement.

Frequently Asked Questions

Does wireless monitoring let us extend or skip the mandated inspection interval?

Not on its own, and not yet everywhere. Under NBIS the two-year visual baseline still stands, though risk-based intervals create room for monitoring data to justify a change on a specific structure. Under MIT 2020 the relationship is being formalised level by level. Treat continuous data as evidence that makes the inspector's visit sharper and better targeted, not as a coupon that buys you out of it.

Can sensors detect corrosion or surface cracking?

Not directly. A network reads structural response, so it can flag that global stiffness has changed and tell you which span to look at, but hidden reinforcement corrosion, chloride ingress and fresh impact damage are surface and material problems that still need an inspector with eyes and a hammer. This is the clearest boundary between the two methods.

Is wireless monitoring accurate enough for modal analysis on a real bridge?

Yes, and it is the measurement we trust most, because it captures the structure under the loads it genuinely experiences rather than under a staged excitation test. Our accelerometers sample up to 640 Hz and feed the MyMove Modal Analysis Tool for frequencies, mode shapes and damping. Forced-excitation campaigns with eccentric mass shakers still have a place for one-off commissioning, but cost and traffic disruption make them impractical for in-service work.

‍