I was at a private collection off Wattletree Road in Malvern a few months back, walking a client through why her Sidney Nolan kept setting off the alarm every time a delivery truck idled outside. The sensor wasn’t faulty. It was calibrated for a frame roughly four times heavier than the one it was actually bolted to. Nobody had touched the threshold since the original hang three years earlier, and in that time the piece had been reframed twice.
So this is the bit nobody tells you when they sell you a vibration sensor: the hardware is only half the job. The other half is a threshold setting matched to the actual mass and mounting of the object, and that number is not universal. A sensor tuned correctly for a 40kg gilt Victorian frame will sit there useless on a lightweight aluminium-framed print, and one tuned for the print will scream every time someone shuts a door two rooms away if it’s bolted to something heavy enough to transmit building vibration through the wall.
What calibration actually means on a vibration sensor #
Most of the sensors we install are either piezoelectric shock sensors or MEMS accelerometer-based units, and both work by measuring acceleration or shock energy against a set threshold, then triggering when that threshold is crossed within a defined time window. The threshold is not a fixed factory number. It’s adjustable, usually via dip switches, a potentiometer, or software in newer networked systems, and it has to be set relative to the mass of what it’s protecting and how rigidly that object is fixed to the wall or plinth.
Heavier objects absorb and dampen shock differently to light ones. A light canvas on a simple D-ring hanger will register a sharp, high-frequency spike if someone tries to lift or pry it, because there’s very little mass to soak up the movement. A heavy framed oil on French cleats transmits a duller, lower-frequency signal across a longer window because the mass and the mounting hardware absorb some of the shock before the sensor ever feels it. If you set both to the same sensitivity, you’ll get false alarms on the light piece and missed events on the heavy one. That’s the whole problem in one sentence, really.
The one-setting-for-the-room mistake #
I’d argue most galleries and a fair few private collectors are running sensors that were never individually calibrated at all, they were set once to a mid-range default across the whole room to save install time, and then left alone. It’s the single most common shortcut I see, and honestly I understand why installers do it, checking each frame individually adds hours to a job. But it means the lightest and heaviest pieces in any given hang are both wrong, just in opposite directions.
This matters more on outdoor and semi-outdoor pieces too, where mounting rigidity and object mass interact with wind load and thermal expansion in ways that indoor sensors were never designed for. We’ve written before about why outdoor sculpture is a security blind spot, and vibration sensing on plinths and bases is a big part of that, because a bronze on a stone plinth behaves nothing like a canvas on drywall.
How to actually test whether your sensor is right #
The proper way to check this is a controlled tap test, done with a known force, not a guess. A technician applies a series of light taps at increasing intensity, checking at what point the sensor triggers, and comparing that against the known weight and mounting method of the specific object. This should be documented per piece, not per room, and it should be redone any time the frame, backing board, or hanging hardware changes.
A few practical checks I’d suggest asking your provider to walk through:
- What weight was this sensor’s threshold calibrated against, and when was that last verified?
- Has the frame been reweighed or reframed since the last calibration was done?
- Is the sensitivity set per object or applied as a room-wide default?
- Does the mounting hardware (cleats, D-rings, security hooks) match what was in place when the threshold was set?
If nobody on your security team can answer the first question with an actual figure, than the calibration probably isn’t real, it’s a factory default that’s been left alone since installation day.
Signs the setting is wrong even if nobody’s told you #
Frequent false alarms on a specific piece, especially triggered by nearby traffic, door slams, or HVAC cycling, usually means the threshold is set too sensitive for that object’s mass. The opposite problem is quieter and more dangerous: a sensor that never triggers at all, even during routine handling by conservators or during a rehang, which suggests the threshold is set well above what the object can actually produce. Neither extreme gets fixed by simply replacing the sensor. It gets fixed by rerunning the calibration against the current weight and mount.
We cover the broader detection side of this in our page on sensors and object monitoring, but the calibration step specifically is where most systems quietly go wrong after the install team leaves.
Reframing and conservation work throws the number out #
Every time a piece goes off to a conservator, or gets a new frame for an exhibition loan, or has its backing board swapped, the mass and mounting rigidity change and the old threshold stops being accurate. This is easy to forget because the sensor itself still looks fine, it’s still on the wall, still reporting to the panel, still showing green on the monitoring dashboard. Nothing about the hardware tells you the number behind it is stale.
This is really an extension of the same documentation problem we’ve written about regarding valuations, where collections drift out of step with their paperwork over time. We covered that in what actually happens if you never revalue your collection, and the same discipline applies to sensor calibration. If you’re updating a valuation after conservation work, that’s the exact moment to also flag a recalibration request to your security provider.
Where the standards actually sit #
There’s no single Australian Standard that says exactly how to calibrate a vibration sensor for frame weight, that level of detail is genuinely down to the installer’s competence and the equipment manufacturer’s guidance. What does exist is the broader intruder alarm system framework under the AS 2201 series published by Standards Australia, which covers design and installation practice for the systems these sensors sit within. For checking an installer’s credentials and accreditation level, ASIAL maintains the industry’s licensing and competency framework, and it’s a reasonable first question to ask any provider quoting on object-level sensing.
Where this touches your insurance conditions #
Some collectors assume a miscalibrated sensor automatically voids a claim if there’s a loss. That’s not how it works under Australian law. The Insurance Contracts Act 1984, specifically section 54, requires an insurer to show that a security failure actually caused or contributed to the specific loss before they can rely on it to reduce or deny a claim, they can’t simply point to a badly set sensor and walk away. That said, we don’t sell or advise on insurance at all, and this isn’t legal guidance, if you want a proper read on how your policy conditions interact with your actual security setup, that’s a conversation for your broker or legal adviser, not us. We’ve touched on the general gap between policy wording and real-world security in the insurance gap nobody checks until it’s too late, which is worth a read alongside this.
What I’d actually check this month #
If you’ve got vibration sensors on frames and you genuinely don’t know when the thresholds were last verified against current frame weight, that’s the thing to chase down before anything else on your security list. It’s cheap to check relative to almost everything else in a collection security budget, and it’s the one thing that quietly drifts wrong without ever throwing an obvious error. Ask for the tap-test documentation by object, not by room. If your provider can’t produce it, that tells you something too.
Anyway, it’s a smaller job than most people expect once someone actually sits down and does it properly, frame by frame rather than room by room.
– Priya Chandran, Security Systems Editor