Summary: A lighting column is a cantilever, and its controlling failure mode is fatigue cracking at the base detail or rotation of a foundation that has quietly lost its grip, not the slow lean that most monitoring specifications are written around. The wind that does the damage is counterintuitive: regulatory inspection guidance records that vortex shedding driven vibration is a known problem for luminaire supports, that it occurs in steady winds above roughly 5 m/s, that it generally stops once wind speed passes about 15 m/s because the flow becomes too turbulent, and that fatigue cracks have initiated and propagated to failure in a single night. Because a healthy pole is designed to sway elastically and return, the usable signal is the return position, not the peak angle. This article sets out the mechanics, the measurement geometry that converts base rotation into tip movement, the sensor selection criteria, and the boundary between what continuous monitoring can and cannot decide.

 

1. The Failure Mode You Are Monitoring For Is Probably Not the One You Wrote Down

Most lamp pole monitoring specifications are written around a slow lean: a threshold in degrees, an alarm when the pole passes it, and a maintenance visit to straighten or replace the column. That model is not wrong, but it describes the least common way a pole reaches the point of collapse, and it encourages a monitoring configuration that is blind to the ways that actually happen.

A lighting column is a cantilever. For a given wind pressure its bending demand at the base rises with the square of height, while its capacity is the section modulus of a hollow tubular shaft that is usually thinnest, and most exposed to water, soil and traffic, precisely where that moment is largest. The structural system is not the pole alone. It is the shaft, the base plate and its welds, the anchor bolts, the concrete foundation, and the surrounding soil, and a chain of failure is only as strong as its weakest link.

The result is that the recorded failure modes cluster at the bottom of the structure, at details a passing inspector cannot see. Regulatory guidance published by the United States Federal Highway Administration for the inspection of structural supports for highway signs, luminaires and traffic signals states that in light poles, cracks are usually observed at the base of the pole or at the weld joining the pole to the base plate or transformer base. Where stiffeners or gussets reinforce the pole to base plate connection, the cracks typically form at the tops of the stiffeners. Where there are hand holes, cracks may appear around the perimeter of those openings. Cracked anchor rods have also occurred.

That list is worth reading twice, because every item on it sits inside the region an operator walks past without looking. A base plate weld is covered by a skirt or by the foundation collar. A hand hole is closed by a cover plate. Anchor rods are inside the base compartment or below the grout line. A gusset top is a detail no one inspects unless they are already suspicious. A monitoring system that only reports a headline angle is not measuring any of this directly; what it can do is detect the change in position that these mechanisms produce, and flag the pole for the detailed inspection that finds them.

2. The Dangerous Wind Is a Steady Breeze, Not a Storm

The intuitive model is that poles fall in storms, so inspection teams are mobilised after a typhoon or a gale and the monitoring system is configured to alert on the largest angles the storm produced. The engineering record points the other way, and the numbers are specific.

When wind flows past a circular shaft, vortices are shed alternately from one side and then the other, producing an alternating force transverse to the wind direction. The shedding has its own periodicity, and when that periodicity locks onto the natural frequency of the pole, the alternating force becomes very large. This is vortex shedding, and the same federal guidance notes that of the support structure types it covers, only luminaires, meaning light poles, have exhibited problems from vortex shedding.

The velocity window is the critical part. According to that guidance, significant vibration does not occur unless the velocity is greater than about 5 m/s, and if the wind velocity is greater than about 15 m/s the wind is generally too turbulent for vortex shedding to occur at all. The dangerous condition is therefore a steady wind in the range of roughly 5 to 15 m/s. The guidance adds two further details that matter operationally. Vortex shedding can lock in and continue as the velocity rises or falls slightly, but if the velocity changes by more than about 20 percent, it will stop. And gusty, variable winds such as those in a severe storm typically will not cause it.

The inversion is uncomfortable. The weather that triggers the inspection is the weather that suppresses the mechanism, and the unremarkable steady afternoon that nobody records is the condition that drives it. It is also worth retiring an old assumption: the same guidance records that vortex shedding can occur in tapered as well as prismatic circular poles of almost any diameter.

Two response modes are usually distinguished. The first, sometimes called harmonic or first mode vibration, is the sway of the pole top under high velocity gusts and loads the base in bending. The second, Aeolian vibration, is the vortex shedding case, is located at or near the middle of the pole rather than the top, and is generally regarded as the more damaging of the two because it is sustained and repetitive. Its visible amplitude near the middle of the shaft can be small enough that a person cannot see it. That is a hard limit on inspection: an inspector has to be at the pole while the right wind is blowing, and even then may need to feel the vibration rather than see it.

The speed at which this can become a structural problem is the argument for continuous measurement. The same inspection guidance describes aluminium light poles that failed after a windstorm in New Jersey, where it was found that the fatigue cracks initiated and propagated to failure in just one night. A pole that can go from intact to collapsed inside a single night cannot be protected by an inspection interval measured in months. It can only be protected by a system that was already watching when that night happened.

One more mechanism is worth separating out, because it is often confused with the above. Galloping is a large amplitude resonant oscillation that occurs on asymmetric members, in other words on mast arms carrying signs, signals or other attachments, rather than on the circular shaft itself, and it is caused by the attachments rather than by the shaft geometry. If your asset includes cantilever arms rather than bare columns, both mechanisms are in scope and they are not the same problem.

3. What a Tilt Sensor Measures on a Cantilever

A tilt sensor measures rotation. What a maintenance engineer cares about is usually a displacement: whether the pole has moved, and by how much at the luminaire. The two are connected by the lever arm, and for the small angles involved the relationship is close to linear. A base rotation of one degree moves the top of a rigid cantilever by approximately the sine of that angle times the height, giving the useful field constant that one millimetre of movement per metre of height is about 0.057 degrees.

Read in the other direction, this conversion is what makes pole monitoring difficult. A permanent change of 20 mm at the head of a 12 m column is only about 0.095 degrees of base rotation. The damage signal you are hunting is a tenth of a degree, which is the same order as the zero point deviation and the temperature drift terms published on many standard industrial tilt sensors. That is the central engineering problem of this application, and it is a signal to noise problem rather than a range problem.

Pole height (base to luminaire)	Tip movement at 0.1 degrees	Tip movement at 0.5 degrees	Tip movement at 1 degree
6 m	About 10 mm	About 52 mm	About 105 mm
8 m	About 14 mm	About 70 mm	About 140 mm
10 m	About 17 mm	About 87 mm	About 175 mm
12 m	About 21 mm	About 105 mm	About 209 mm
15 m	About 26 mm	About 131 mm	About 262 mm

The table also shows why range and resolution have to be chosen together. A pole in a gust can swing several degrees at the base before anything is wrong with it, so the sensor has to remain within range across that normal elastic sway. At the same time the permanent set that is the real signal is a small fraction of a degree. Choosing a range so tight that the sensor clips during the first storm loses the data exactly when it is most informative, and choosing a range so wide that the resolution at the small end is coarse loses the damage signal instead. Both mistakes are common and only one of them produces an obvious error message.

A second consequence of the geometry is that one sensor at one height measures one thing: the rotation of the pole at that point. It cannot by itself tell you which part of the structure produced that rotation. Two sensors, or one sensor combined with a considered choice of location, change what the data can answer.

Measurement arrangement	What the data shows	What it cannot resolve
One sensor on the shaft just above the base flange	The combined rotation of the foundation and the lower shaft, which is where permanent change usually appears first	Whether the rotation came from a moving foundation or from bending of the shaft above it
A pair of sensors at two heights, for example base and mid-height	The component common to both readings is rigid body rotation; the difference between them is the bending of the shaft over that interval	Where a crack is, or whether a specific weld detail has failed; it identifies the active mechanism, not the defect location
One sensor high on the pole, near the top	The largest amplitude motion, because tip movement is amplified by the lever arm	A healthy pole returns to its starting point after swaying, so the only usable signal is the small residual, and it is dominated by every source of elastic movement

4. Foundation Rotation and Shaft Bending Are Different Problems

Two mechanisms produce almost the same reading on a single sensor and require completely different remedies, so the distinction is worth building into the monitoring design rather than discovering later.

If the foundation has rotated, the whole assembly tilts as a rigid body. Every point on the shaft rotates by the same amount, the shaft remains straight, and the correct response is geotechnical: investigate the soil, the groundwater, the concrete and the anchor group. If the shaft has begun to yield or a base detail is failing, the angular change is concentrated near the base and the intervals above it stay close to zero. The correct response is structural. If both are happening, and on an old pole in a coastal city that is a realistic scenario, a single sensor cannot separate them, while a two height pair can.

The anchor group deserves its own attention because it sits between the two mechanisms. Anchor bolts carry tension, shear and long term loosening, and their behaviour is not simply elastic. A group that has lost preload or has begun to corrode allows the base plate to rotate on the concrete, which reads as a foundation problem on the sensor even though the shaft is undamaged. Conversely, a foundation that is rotating steadily will progressively load one side of the anchor group in tension and can eventually crack the rods themselves, which is one of the failure modes named in the inspection guidance.

Corrosion is the mechanism that makes all of this time dependent, and it is concentrated in the least accessible zone. The European standard for steel lighting columns, EN 40-5, divides the column into an exterior area, a ground section and an interior for the purpose of specifying corrosion protection, a division that reflects operational reality: the ground section, at and just below grade, is simultaneously the most aggressive environment and the hardest part of the structure to examine. Protective coatings such as hot dip galvanizing work until they are breached, and after that the loss of wall thickness reduces the section modulus directly, at exactly the elevation where bending demand is highest. A pole can therefore lose a significant fraction of its capacity without changing its plumb at all, which is why a monitoring programme that only watches angle is a partial view and why periodic physical inspection remains mandatory.

5. Why the Return Position Is the Signal

A healthy lighting column is designed to move. Structural verification under the EN 40 series covers resistance to horizontal wind loads, through the design and verification parts dealing with calculation and with testing, alongside a separate assessment of passive safety under vehicle impact under EN 12767. A pole that bends and recovers under wind is doing what it was designed to do, and a monitoring system that alarms on elastic movement will produce a stream of alerts that correspond to nothing wrong.

The usable signal is therefore the position the pole comes back to. Elastic movement is reversible; damage is not. What matters is the difference between the rest position before an event and the rest position after, which is why the recommended measurement is not the peak angle observed during a gust but the median of the sensor output over a quiet window, ideally at night when thermal and traffic effects are at their minimum.

This has a direct consequence for how a monitoring system should be configured on a pole. The alarm logic needs a rest position comparator, not a level detector, and it needs a persistence requirement so that an instantaneous gust does not trigger anything. A useful discipline is to treat the peak angle as diagnostic information about loading, and the change in rest position as the maintenance trigger.

There is one competing signal to separate out at the same time. A pole is an outdoor structure with one face in the sun and one in shadow, and the resulting daily thermal cycle produces a repeating movement that reverses with the sun. A genuine structural change accumulates and does not reverse. If the apparent movement repeats every day and changes sign between morning and afternoon, it is thermal. If it appears once and stays, it is structural. That test is free and it is the first one to apply.

6. Smart Poles Have Changed the Load Case

The lighting column as a structural type was developed for a luminaire, and its design assumptions reflect that: one relatively light fixture at a defined height on a defined arm. A smart pole, or a multi service pole that consolidates lighting, cameras, communications equipment, sensors, signage and displays onto a single mast, is a different structure even when the shaft looks familiar.

Every added device contributes projected area at a moment arm, so it increases both the total wind force and the base bending moment. Devices are also frequently mounted at six to nine metres, where the lever arm is long. Industry guidance for smart solar lighting systems suggests that for an eight to ten metre multi-function pole, the extra exposed area of mounted equipment can reach roughly 0.3 to 1.2 square metres and raise the base moment by 20 to 50 percent depending on bracket geometry and display size. Those figures are an industry estimate rather than a universal rule, and the actual value depends entirely on what is fitted, but the direction is not in dispute.

The procurement failure that follows is structural rather than commercial. The pole, the arm, the brackets and the foundation are one load path, but they are frequently specified as separate line items by different parties, and a pole rated for a given wind speed with a luminaire is not automatically compliant once additional equipment is added to it. In Chinese practice, the consolidated multi service pole is explicitly treated as a structure with a more complex load case than an ordinary street light, because of its long cantilevers and heavy mounted equipment, and it is subject to regular safety inspection with commonly quoted intervals of one to three years, plus a dedicated special inspection after severe wind, typhoon or earthquake events. Inspection practice of that kind reflects exactly the point made in this article: the failure modes concentrate in hidden details and need physical examination.

For monitoring, the practical implication is that the baseline is not permanent. A tilt baseline established at commissioning describes a pole carrying the equipment it had on that day. If cameras, antennas or displays are added later, the load case changes and the pole may adopt a new rest position without anything being wrong. A monitoring programme that does not record when the pole’s configuration changed will eventually generate an alarm that is really a record of a change in payload.

7. Choosing a Sensor for a Pole

This application has an unusual advantage and an unusual difficulty. The advantage is that a lighting column normally has a permanent electrical supply inside the base compartment, which most civil monitoring sites do not, so the usual battery and power budget constraints can be relaxed. The difficulty is the signal to noise problem described above: the quantity of interest is small, and the environment in which it must be measured is thermally and mechanically noisy.

Requirement	Why it matters on a pole	What to look for	Common mistake
Range wide enough for elastic sway	A pole can swing several degrees at the base in a gust with nothing wrong with it, and the sensor must stay in range throughout	Range matched to the observed sway at the specific site, checked against the tip displacement table above	Selecting the tightest available range to gain resolution, then clipping the first time a storm arrives
Repeatability at the small end	The permanent set that matters is a small fraction of a degree, so the number that governs is how faithfully the sensor reproduces the same position	A repeatability figure published separately from the accuracy and resolution figures	Reading the resolution figure and assuming the sensor can resolve changes at that size in the field
Published temperature behaviour	An outdoor pole sees a wide daily and seasonal temperature swing, and a drift term generates a daily cycle that looks exactly like structural movement	A stated zero point temperature drift, and a datasheet note describing the conditions under which the accuracy is quoted	Assuming the headline accuracy applies uniformly across the full operating temperature range
Ingress protection matched to the site	Rain, spray, dust, salt and ultraviolet exposure for a decade or more, with the sensor mounted low where water collects	An IP rating appropriate to a permanently wet or coastal location, and a cable gland or connector rated at least as highly as the housing	An IP67 sensor housing defeated by an unprotected cable entry or an unsealed junction box
A rigid, defined mounting reference	The pole surface is curved, often tapered and often galvanized, and the sensor must move exactly as the structure moves	A machined mounting face or a bracket that makes contact on a defined plane and is positively fixed	Clamping to a tapered shaft through a soft pad or a single hose clip, which allows slow creep that is indistinguishable from real movement
Power and communication route	There is usually a supply and often a control cabinet already at the pole, so a wired sensor can be a simpler answer than a wireless one	RS485 or an analogue loop back to the cabinet where wiring is permitted, or a cellular node where it is not	Defaulting to a battery powered wireless node when mains power is already available in the same base compartment

8. Baseline, Alarm Logic and the Limits of What Monitoring Can Decide

Setting up a pole monitoring system is mostly a matter of discipline about the baseline. The rest position must be recorded in still air and after the sensor has thermally settled, because a baseline captured during a windy, sunny afternoon encodes the weather of that afternoon into the reference against which every future reading will be compared. The pole height, the equipment fitted, the mounting arrangement, the sensor serial number and the date should be recorded with it, so that later additions of equipment can be recognised as a change in configuration rather than misread as structural movement.

Alarm logic on a pole should follow a sequence rather than a single number. The first test is whether an apparent movement reverses daily, which separates thermal effects from real change. The second is whether a change persists and is confirmed by a repeating pattern of rest position measurements. The third is whether the change is large enough relative to the sensor’s own repeatability and drift budget to be meaningful at all. A threshold set below that budget produces regular alerts that correspond to nothing, and a stream of alerts that operators learn to ignore is worse than no system, because it consumes the attention that the one real event will need.

It is equally important to be explicit about the boundary. Continuous tilt monitoring on a pole can tell you that a specific pole has changed its rest position, when the change started, and whether the movement is continuing. It cannot tell you whether a particular weld has cracked, how much wall thickness has been lost to corrosion, whether an anchor group has lost preload, or whether a given pole is safe to leave in service. Those are findings that require a physical inspection by a competent person, and in most jurisdictions the periodic inspection of street lighting and road equipment support structures is a regulatory obligation that monitoring does not discharge.

What monitoring changes is how the inspection effort is spent. A city with tens of thousands of poles and a limited budget for climbing inspections has to choose. A monitoring programme converts that choice from a schedule driven by the calendar into one driven by evidence, directing attention to the poles that have measurably changed and to the period in which the change occurred, while leaving the statutory inspection regime, the assessment of structural adequacy and every repair decision exactly where they belong.

Frequently Asked Questions

1. How much tilt is dangerous for a lamp pole? There is no universal number, because the answer depends on pole height, shaft section, wall thickness, foundation type and age. What is more useful is the change relative to the pole’s own rest position: a column that has always stood slightly off plumb and is not moving is a different situation from one that has shifted by 0.1 degrees in the last month. For scale, a base rotation of one degree on a ten metre pole corresponds to roughly 175 mm of movement at the luminaire, which is a lean you can see, while the fatigue mechanisms described in this article can progress at the base with no visible lean at all. The decision on whether a measured change is acceptable belongs to a qualified structural engineer, not to the sensor.
2. Why is vortex shedding more dangerous than a storm? Because it is steady, repeatable and capable of locking onto the natural frequency of the pole, and because it occurs in a wind speed band that no one treats as a hazard. Published inspection guidance records that significant vibration from vortex shedding does not occur below roughly 5 m/s, that winds above roughly 15 m/s are generally too turbulent for the mechanism to operate, and that gusty variable winds in a severe storm typically will not cause it. The damaging condition is therefore an ordinary steady wind of about 5 to 15 m/s, and the same guidance records light poles where fatigue cracks initiated and propagated to failure in a single night.
3. Can tilt monitoring replace periodic pole inspection? No. Inspection of road equipment support structures is a regulatory obligation in most jurisdictions, and a monitoring system cannot judge a weld, measure remaining wall thickness or evaluate the condition of a foundation and anchor group. What continuous monitoring does is change how a limited inspection budget is allocated, by identifying which poles have changed their rest position, when the change occurred and whether it is continuing, so that physical inspection is directed by evidence rather than by a fixed calendar.
4. Where should the sensor be mounted on a lamp pole? The most informative single position is on the shaft immediately above the base flange, on a rigid bracket with a defined contact plane, so that the sensor records the rotation of the pole relative to the foundation. That is where permanent change typically appears first. A second sensor at mid-height allows the rigid body rotation of the assembly to be separated from bending of the shaft, which decides whether the correct response is geotechnical or structural. Avoid the door or hand hole region for mounting, since the opening is itself a stress concentration and cracks are known to form around its perimeter.

---