Hybrid systems: point sensors and DFOS fibre-optic measurements

In brief

• Point sensors (vibrating wire, thermistors, accelerometers, inclinometers) measure automatically, every 15 minutes by default, or more often, because that is how the logger is set up - and they provide trends, thresholds and notifications.
• DFOS (distributed fibre optic sensing, distributed fibre optic measurement) measures strain along the entire length of the sensor, but usually periodically: with an interrogator (optical logger) brought in by the measurement crew.
• Division of roles: point sensors watch the time axis, fibre optic watches the length; an alarm from the point sensors can trigger an additional DFOS session.
• Five arguments for a hybrid system: knowledge, versatility, reliability (two independent techniques), flexibility, savings - without buying an interrogator and with fewer expensive point sensors.
• The limitations are real: periodic DFOS does not alarm on its own, requires crew access, and cannot replace dynamics. It is cheapest to install it during construction - after concreting, this option becomes more expensive.

Two questions that one sensor type cannot answer

Anyone who has designed monitoring for a bridge or a large hall knows the dilemma. You have a budget for a dozen sensors, the structure is several hundred metres long, and a crack does not ask the designer where it is allowed to appear. The asset owner asks two questions that sound similar, yet require completely different tools.

First: what is happening to the structure over time? Is the span deflection increasing year by year, or is it only “breathing” with temperature? Did anything change last night? Continuous time measurement answers this - a sensor that records every quarter-hour by default, or more often, year after year, in the same location.

Second: where exactly is something happening? Has a crack appeared between control sections, unseen by anyone because it is inside a box girder? Where do stresses concentrate after an overload? Continuous measurement along the length answers this - a sensor that “sees” every centimetre of the element, even if only once every few months.

A classic point system answers the first question well and the second badly: the space between measurement points is blind. A classic fibre-optic measurement answers the second question brilliantly and the first weakly: the periods between measurement sessions are also blind, because the interrogator comes only from time to time. A 2025 industry publication on hybrid systems using bridges as an example puts it briefly: the limitation of point methods is their locality, and the limitation of fibre-optic methods is periodicity, which comes from economics, not from the technology itself. A hybrid system is created by combining these two gaps into one advantage.

If you are still shaping your understanding of what structural monitoring is and how it differs from inspection, start with the complete SHM guide. Here we focus on one slice: how to combine two techniques so you do not pay twice for the same information.

Point sensors: automation and early warning

A point sensor is a transducer that measures one physical quantity (strain, temperature, inclination, acceleration) at one selected location on the structure, permanently connected to a logger. In long-term monitoring, these are most often vibrating wire sensors for strain, force and pressure, thermistors, MEMS accelerometers and inclinometers. Why vibrating wire sensors in particular, and why they are still the standard, is described separately in the text about vibrating wire sensors.

The basic requirement for an early warning system is simple: the measurement must be automatic and continuous over time. The logger collects readings, sends them to the server, and the data are compared with thresholds and analysed for trends. This gives three things that cannot be obtained otherwise:

• trend - you can see whether the value in a given section is rising, falling or oscillating around a stable level, and whether that oscillation matches temperature;
• forecast - with several years of records, you can assess where the trend is heading;
• alarm - exceeding a warning or alarm threshold sends a notification to authorised people within minutes, not at the next inspection.

This has a cost, which the same publication states plainly. First, the measurement exists only where the sensor is installed; selecting the “right” sections is not trivial, and even then nobody knows what is happening between them. Second, the unit cost of a measurement point is relatively high, because long-term monitoring requires high-grade sensors. Third, every sensor needs a cable to the logger, and on a large asset the cable routes can be long and awkward. The cost of the loggers themselves is usually negligible at this scale.

That locality is not a defect of a particular manufacturer. It is a feature of the approach. And fibre optic is what fills that gap.

DFOS: geometrically continuous measurement and crack detection

DFOS (distributed fibre optic sensing) is a measurement technique in which the fibre optic itself is the sensor: the interrogator sends light into the fibre and, based on the backscattered signal, determines strain or temperature along the full length of the fibre. There are no discrete “sensors at points” here. There is one continuous measurement line - from a few centimetres to hundreds of metres on a single asset, and in linear infrastructure even kilometres.

Monolithic sensors laid along the element

In structural applications, a bare telecommunications fibre is not laid. Monolithic sensors are used - in the literature and on the market, these include EpsilonSensor-type sensors with a composite core, a diameter of several millimetres, intended for embedding in concrete or ground, or for surface mounting. The fibre is integrated into the matrix, and the outer roughened sheath ensures bond to concrete or mortar. Such a sensor is laid along the element: in a near-surface groove filled with installation mortar, by bonding to the concrete surface, or - preferably - directly in the concrete during construction, with the reinforcement. From that point on, the element has a “nervous system” along its full length.

Periodic measurement, by crew, with an interrogator

In most structural applications, DFOS is not permanently connected. The measurement crew arrives with the interrogator, connects to the sensor ends brought out, performs a session and leaves. The result is a strain profile along the element at a given moment, compared with the reference session (the “zero” measurement) and with previous sessions. The reason is economic, not technical: interrogators are expensive, so one device serves many assets. A 2025 industry publication identifies this directly as the main brake on the wider adoption of DFOS.

Spatial resolution - why it matters

The spatial resolution of DFOS is the smallest section of fibre for which the interrogator determines a separate strain value; the smaller it is, the denser the profile and the more reliable the detection of local phenomena. It depends on the measurement physics. A literature review by Barrias, Casas, Villalba, Sensors 2016, states that Rayleigh-based techniques (OFDR) provide millimetre-level resolution, but over lengths up to about 70 m, while Brillouin-based techniques (BOTDR/BOTDA) provide about 1 m, but over tens of kilometres. On a bridge, the first class matters; on a gas pipeline, the second.

In structural practice, we are interested in millimetre or centimetre resolution, because that is the “width” of the strain disturbance zone around a crack. For scale: in a documented deployment on a prestressed bridge, the authors adopted an interrogator resolution of 5 mm, which for more than 1 000 m of sensor gave about 208 000 measurement points in a single session (2025 industry publication). You cannot install that many point sensors, neither technically nor economically - and that is the whole point.

What does this mean for diagnostics? The strain profile of a healthy concrete element is smooth, with minor fluctuations caused by material heterogeneity. A crack gives a sharp local peak on the profile, and stress concentration gives a clear anomaly. You do not have to guess where to place the sensor: the extreme cannot hide between points.

How to combine them: division of roles and trigger

A hybrid system is not “a bit of one and a bit of the other”. It is a division of roles in which each technique does what it does well, and does not pretend to be the other.

Point sensors are the system’s “on-duty staff”: they work every 15 minutes by default, or more often, day and night, and say whether anything has changed. DFOS is the “expert on call”: it comes periodically, checks the full length of the element and says where the problem is - or confirms that there is none.

Trigger: from alarm to DFOS session

A trigger in a hybrid system is the rule under which an event recorded by automatic point sensors - threshold exceedance, unusual trend, vibration event - launches an additional, unscheduled DFOS measurement session.

The mechanism is simple and should be written into the monitoring procedure:

1. The monitoring plan defines the DFOS session schedule (for example, a reference session after installation, then cyclically - illustratively once a year or after each winter).
2. Point sensors have WARNING and ALARM thresholds on each channel and are observed for trend after temperature influence has been removed. How to set thresholds so they do not call in false alarms is described in the text on warning and alarm thresholds.
3. When an ALARM appears, a sustained WARNING after temperature compensation, an exceptional event (overload, impact, flood) or a trend that temperature does not explain - the procedure provides for ordering a DFOS session outside the schedule.
4. The crew measures the full length of the sensor, and the result is compared with the last regular session and with the trends from the point sensors. The answer is: “crack at km X+YYY, open since the last session” or “profile smooth, point change has another explanation”.

Illustrative scenario: in the mid-span section, strain after temperature compensation has been rising for the third month in a row, although in previous years it returned to zero. No channel has exceeded the alarm threshold, but the trend is there. The DFOS session shows a local peak a few metres away from the control section - where there was no point sensor. Without the hybrid approach, you would learn about it at inspection, if the crack were visible from the outside.

Five arguments for a hybrid system

Industry literature lists the benefits of hybrids as a set. We group them into five arguments that come up in conversations with investors and designers, and add the rationale.

1. Knowledge: time and length at once. Point sensors give the history over time, DFOS gives the picture in space. One system without the other answers only half of the asset owner’s questions. A hybrid answers both halves and the combined question: “does what I see in the trend have a cause at a specific location?”.

2. Versatility: the same techniques, different assets and life-cycle phases. Fibre optic suits long elements (girders, spans, collectors, embankments), point sensors suit locations where you need a number every 15 minutes by default, or more often (piers, critical sections, prestressing zones). DFOS measures strain, temperature, displacement and vibration; point sensors cover strain, force, inclination and vibration. The set is designed around the asset, rather than forcing the asset to fit one technique.

3. Reliability: two independent techniques provide redundancy. A vibrating wire sensor and a fibre optic sensor have different measurement physics, different signal paths and different sources of error. If both techniques show the same thing in the same section, you have confirmation. If they differ, you have a signal to check the installation, compensation or the structure itself before anyone makes a decision based on one reading. In an expert assessment, two consistent, independent techniques carry more weight than one.

4. Flexibility: at any stage and at any rhythm. Fibre-optic sensors can be embedded in concrete during construction or added to an existing asset in a groove or on the surface. Point sensors are added when there is a need for continuous supervision of a specific section. You densify the DFOS session schedule after an alarm (trigger) and space it out again when several consecutive sessions confirm stability.

5. Savings: no amounts, but the logic is clear. We will not give monetary figures here - every asset is priced separately. The mechanism, however, is repeatable:

• the fibre-optic sensor itself is cheap per metre compared with an equivalent number of point sensors;
• you do not buy an interrogator - DFOS measurement is ordered as a service, and one device serves many assets;
• fewer expensive point sensors and less cabling, because they do not have to “imitate” continuous measurement along the length;
• deferred investment - fibre optic installed during construction can wait years for the first measurement, and you incur the cost of a session only when you need it;
• life-cycle cost - the hybrid splits spending into installation (one-off), the automatic part (ongoing maintenance and platform cost), and DFOS sessions (on demand), instead of concentrating it in one purchase.

For the budget, that means this: instead of one large line item called “monitoring system”, you have three items of different character - and only one of them (the automatic part) is a recurring annual cost. How to calculate this for your own asset and what goes into five-year TCO is described in the text on monitoring system cost and TCO.

Which assets

A hybrid makes sense where the structure is long or extensive, and the asset owner wants both early warning and a periodic full picture. Literature points to several groups:

• hydrotechnical and geotechnical assets - dams, levees, embankments, where fibre optic covers the full length of the crest or slope, and point sensors (piezometers, inclinometers) watch selected sections;
• linear infrastructure - gas pipelines, railway lines, tunnels, collectors, where the length itself excludes dense point measurement;
• large-area assets - stadiums, halls, roofs with large spans, where the girders are long and snow load acts over the whole surface;
• industrial assets - power plants, warehouses, structures under equipment;
• bridges - prestressed, extradosed, pedestrian bridges; here hybrids are best documented and most often implemented.

The common denominator: an element length that cannot reasonably be covered with point sensors, plus several sections in which a number is needed every 15 minutes by default, or more often. What to measure on a bridge asset and how to read the results is described in the bridge monitoring guide for asset owners.

Examples described in the literature

These are not our implementations. The cases below come from 2025 industry publications on hybrid systems; we cite them without asset names, operators or locations - what matters is the system arrangement and the conclusions.

Prestressed bridge, five spans, over 260 m

Asset: a road bridge with a load-bearing structure in the form of a prestressed reinforced-concrete box girder, total length over 260 m. Due to doubts about durability, the existing asset was equipped with a hybrid system.

Fibre-optic part: monolithic sensors in four measurement lines along the full bridge length, inside the box girder, in near-surface grooves filled with installation mortar - in total more than 1 000 m of fibre-optic sensor for periodic measurements of strain distribution and crack detection; additionally, short sections bonded directly to the concrete along the prestressing cables.

Automatic part: 12 vibrating wire strain sensors with reference thermistors, arranged in three control sections with four transducers each.

Findings after the first year (according to the publication): strain histories over time showed full dependence on temperature changes on a yearly basis, with no worrying trends. DFOS profiles from three sessions were smooth, with no local extremes typical of cracking - no open cracks were found along the entire bridge length, which was the most important information for durability assessment. One more observation by the authors: one session was carried out at a temperature about 30 °C lower than the reference session and the profile shifted into the negative range (thermal contraction); a session performed under conditions similar to the reference gave a profile close to zero. For this reason, thermal conditions during sessions must be recorded, and continuous temperature measurement from the automatic part is essential here.

Extradosed bridge

The second arrangement described in the literature concerns an extradosed bridge. EpsilonSensor-type fibre-optic sensors were laid along the entire span, and the automatic part was built from vibrating wire sensors (strain, rotation), thermistors and accelerometers. A camera was added to the system to correlate truck passes with measurement data - a peak on the chart can be assigned to a specific event, rather than guessed to be an overload or an artefact. This example shows the versatility of the hybrid: statics (vibrating wire, DFOS), dynamics (accelerometers) and context (camera) in one system.

What this means for you: both arrangements have the same skeleton - fibre optic along the length, point sensors in sections, temperature as background - and you can transfer such a skeleton to your own asset, changing only the proportions.

Limitations we discuss at the first meeting

A hybrid is not a cure-all, and there are situations where we would not propose it.

Periodic DFOS does not alarm on its own. Between sessions the fibre optic is silent. If someone sells you “fibre optic monitoring” without the automatic part and without a permanently installed interrogator, you are buying periodic diagnostics, not an early warning system. That may be fine, as long as you call it what it is. If you need a night-time alarm, the automatic part is mandatory.

It requires crew access. A DFOS session means people travelling with expensive equipment, access to the sensor ends (sometimes inside a box girder, at height, in a tunnel) and time for the measurement. Where access is difficult, sessions will be less frequent - this must be reflected in the plan.

Installation during construction is the cheapest and best option. Embedding the sensor in concrete with the reinforcement gives the best bond and the lowest cost. On an existing asset this can still be done (groove, bonding), but more expensively and with location constraints. If you have influence on the design, decide on the fibre optic before concreting, not after.

Dynamics belongs to point sensors. Vibrations, vibration events, spectral analysis, response to pass-by - these are measured by accelerometers in continuous mode. In a hybrid, fibre optic complements statics, not dynamics.

When it makes sense, when it does not - the decision in one table

The most common concern we hear: “I will pay for fibre optic that nobody will ever measure later” or, on the other hand, “I will buy monitoring and nobody will call at night anyway”. Both concerns are valid, and both are solved by matching the system to the situation, not to the catalogue.

If your asset falls into the first or second row, the decision has a deadline: concreting or the next repair window, because that is when fibre optic is cheapest to install. If you are in one of the last three rows, you will save money by not buying the hybrid.

What it looks like at Inclify

Inclify handles the automatic part of the hybrid system: point sensors - vibrating wire sensors, thermistors, accelerometers, inclinometers - connected through loggers or existing hubs, with readings every 15 minutes by default, or more often and online data. On each channel you set WARNING and ALARM thresholds with hysteresis, SMS and e-mail notifications go to authorised people, and the case has a history - who acknowledged it, who muted it and for how long. Strain from a control section is plotted on one multi-axis chart with temperature, so that the annual “breathing” of the structure can be distinguished from a permanent trend - exactly as in the prestressed bridge case described above.

The decision on the trigger is made by a human, and the platform gives them the data needed for it: charts and time comparisons, a 7-day and 30-day risk assessment report, and anomaly detection (deviation from the baseline window) in AI analysis. The platform does not have a “trend alarm” - you assess the trend on the chart and in the report, while the threshold alarm rings when the value exceeds the set level. Results from periodic DFOS sessions are compared with trends from the point sensors: you receive the DFOS profile from the measurement crew as the session result and read it alongside the charts from the automatic part; the platform does not store or display fibre-optic profiles. How the sensors and loggers are connected is described on the platform and device integration page, and an example bridge arrangement is on the bridge monitoring in Inclify page.

Sources and further reading

• “Hybrid systems - introduction using bridges as an example” - 2025 industry publication on combining automatic point sensors with periodic DFOS measurements: PDF on shmsystem.pl
• “Hybrid systems - case study: implementation on a prestressed bridge” - 2025 industry publication describing the arrangement and first year of operation of a hybrid system on a five-span prestressed bridge: PDF on shmsystem.pl
• “Hybrid systems - case study: gas pipeline” - 2025 industry publication on a hybrid system in linear infrastructure: PDF on shmsystem.pl
• Barrias A., Casas J.R., Villalba S., “A Review of Distributed Optical Fiber Sensors for Civil Engineering Applications”, Sensors 16(5):748, 2016 - review of DFOS techniques, spatial resolution and ranges: pmc.ncbi.nlm.nih.gov
• EpsilonSensor data sheet (monolithic DFOS strain sensor) - Nerve-Sensors: nerve-sensors.com
• Technical documentation for vibrating wire sensors and thermistors for long-term monitoring - geokon.com
• Inclify - Structural monitoring (SHM): complete guide
• Inclify - Vibrating wire sensors: why they are still the standard
• Inclify - Bridge and viaduct monitoring: what to measure, how to interpret, when to react

What next

If you are wondering whether your asset - a bridge, hall, dam, collector - is best served by point sensors only, fibre optic only, or a hybrid, let us talk about system selection. The Inclify team has been designing and maintaining measurement systems for 15 years, from stadiums to hydrotechnical sluices, and can tell you when fibre optic is not needed. In a turnkey model, we select the sensors, install them and bring the automatic part online in the platform; if you already have sensors and loggers, integration takes a few days.

One thing has a deadline: if the asset is in design or before concreting, the fibre-optic decision is made now - later, the same installation requires grooves and bonding, which costs more and covers less. Book a system selection consultation - describe the asset and stage, and we will reply within 24 hours.