Product Deep Dive: Everything You Need to Know About Kaiterra’s New Sound Sensor

Kaiterra just launched its first-ever sound sensor, bringing continuous sound monitoring to the same platform that already tracks air quality, thermal comfort, and light. It measures how loud each space is, how much its sound level varies, and what frequencies make up that sound, and it's built with four layers of security so no audio ever leaves the sensor.

Kaiterra’s sound sensor gives workplace leaders and building managers a way to finally measure and resolve the number one workplace complaint, noise, and deliver acoustic comfort so every occupant can feel well and do their best work.

Why Should Workplaces Monitor Sound Continuously?

Noise is consistently the most reported cause of dissatisfaction in the office, with measurable impacts on well-being and productivity, yet most workplaces aren’t measuring it day to day. Many rely only on occupant complaints, which often surface months after a problem starts, or on one-off assessments like a handheld sound meter or an acoustic survey.

While useful, those methods only capture a single point in time, so they inevitably miss how a space’s sound varies as occupancy, utilization, building systems, and outdoor noise change throughout the day. Continuous sound monitoring fills that gap by showing you where noise problems occur, what’s causing them, and whether each space delivers acoustic comfort for the people working in it, which is exactly what Kaiterra’s new sound sensor was built to do.

How Does Kaiterra's Sound Sensor Help Deliver Acoustic Comfort?

Kaiterra's sound sensor continuously measures the level, variability, and frequency of sound in occupied spaces, giving workplace and facilities teams the data they need to match each space to the people and work in it and deliver acoustic comfort. After a decade of measuring air quality, thermal comfort, and light, Kaiterra is completing the picture of indoor environmental quality with sound.

What Makes Kaiterra's Sound Sensor Different From Other Methods of Measuring Office Noise?

Here's what Kaiterra's sound sensor gives you that an acoustic survey, sound meter, or occupant complaints can't:

• It measures continuously. Acoustic surveys and sound level meters capture a moment or a day. Kaiterra's sensor samples sound every second of every day, building hundreds of hours of data per room each year while people are working.
• It measures more than decibels. It tracks level, variability, and frequency together, and identifies the likely source of the sound, so you can see why a space feels the way it does, and know how to fix any issues.
• It puts sound on one platform with the rest of IEQ. Sound sits alongside air quality, thermal comfort, light, and occupancy in Kaiterra's all-in-one IEQ platform, so teams can see how these factors interact.
• It's private by design. Four layers of security, enforced in hardware and reviewed by an independent security firm, make sure it can't be turned into a recording device.

What Does Kaiterra's Sound Sensor Measure?

Kaiterra's sound sensor monitors all three characteristics of acoustic comfort:

By analyzing these data points side by side, the Kaiterra Data Platform pieces together a clear acoustic picture of each of your spaces, so you know which spaces have acoustic issues, where those issues come from, and how to fix them.

What Insights Does Kaiterra's Sound Data Provide?

For high-level insights and decision-making, the Kaiterra Data Platform plots level against variability, so you can quickly see which acoustic profile a space has.

Variability often matters more than level, and low variability is the target for almost every space. A café can deliver acoustic comfort with a loud yet steady environment, while a library can fail to deliver it with quiet yet spiky sound.

How Can Workplace Teams Use Kaiterra's Sound Data?

Continuous sound data geared toward acoustic comfort is designed to help you resolve acoustic issues faster and at a lower cost, reduce occupant complaints (even before they arise), and make your office a comfortable, productive place where your team actually wants to be.

Continuous sound monitoring turns noise from a vague complaint into something you can locate, explain, and fix. 

-- Liam Bates, CEO of Kaiterra

A few examples of how Kaiterra's sound data can help you achieve a better acoustic environment:

• Check whether acoustic zones work as designed. Continuous data shows whether a focus zone still supports focus at 3 p.m. on a Wednesday, or whether a collaboration space is consistently lively or simply too loud. It confirms your design is still doing its job long after fit-out, making the office worth the commute and delivering a better employee experience from the space you already pay for.
• Diagnose the cause of a noise problem. A distracting open-plan office, a meeting room where conversations carry, and a focus zone with a persistent hum all require a side-by-side analysis of sound level, variability, and frequency to diagnose. Continuous data makes that analysis possible, showing when and where each problem occurs and what's driving it.
• Spend on the right fix, then confirm it worked. Instead of blanket efforts to reduce office noise, like installing panels everywhere or doing a costly renovation, continuous data narrows the cause to a single HVAC unit, zone boundary, or team on speakerphone, so you resolve issues faster and at lower cost. Before-and-after data then shows whether the change held, or if more treatment is needed.
• Give occupants a choice of where to work. Sharing acoustic data allows employees to pick the space that best suits their preferences or current task, from a quiet corner for deep work to a livelier area for collaboration. When people have that choice, fewer end up stuck somewhere that doesn't suit their work, which reduces complaints.

How Does Kaiterra's Sound Sensor Protect Your Privacy?

Kaiterra's sound sensor ensures complete privacy with four layers of hardware-enforced security that prevent it from capturing intelligible speech or sending audio anywhere.

That protection matters because every sound sensor contains a microphone, since that's how sound is measured. A microphone is a transducer, a device that changes one form of energy into another. Sound arrives as tiny variations in air pressure; those variations move a thin diaphragm, and the microphone turns that movement into an electrical signal the sensor can measure.

For anyone considering installing a microphone in an office, meeting room, or classroom, the number one question is often about privacy, and rightfully so. How do you know the sensor isn't recording all of your conversations and sending them to the cloud? And how do you know that audio is safe from hacking?

We built our sound sensor with this exact concern top of mind, which is why we decided to bypass the question of recorded audio completely. The sound sensor is a self-contained module that plugs into the Sensedge Go (Kaiterra's wireless, battery-powered monitor).

It sits inside the monitor's case but is physically separate from the rest of the device, with its own microphone and its own microcontroller unit (MCU), a tiny computer chip that handles all of the audio processing. Because that processing happens on the sound sensor itself, audio never reaches the monitor, the network, or the cloud.

This design fits into our four-layer security architecture, which ensures the conversations in your office stay private.

Layer 1: The Microphone Barely Listens at All

Saying the word “microphone” out loud takes about a second. In that same second, Kaiterra's microphone powers up, measures the audio for 87 milliseconds, then powers down for the rest of the second. That's about a twelfth of a second, a bit faster than the blink of an eye: enough time to measure the sound in a room, but not enough to make out any words.

No audio is scrambled or filtered; each sampling window is simply too short to capture intelligible speech. To prove it, we had a group of people put on headphones, listen to 36 sentences cut to different lengths, and write down every word they could make out. The shorter the clip, the less they caught:

At 0.4%, listeners did no better than guessing. To verify these results, our engineers ran the same test on an AI model. The model was trained on 28 hours of audio, cut into the same 87-millisecond segments the sensor hears. Out of 36,000 words, it recovered only 8, a 0.02% success rate (less successful than randomly guessing words from a dictionary).

Layer 2: The Microphone Can't Be Configured to Listen Longer

The 87-millisecond window isn't a setting that can be changed, because it's hardwired:

• On the sensor's circuit board, a timer chip pulses once each second.
• Each pulse powers a simple circuit, made of a resistor and a capacitor, that switches on the microphone's power for 87 milliseconds and then switches it off again.
• The length of that pulse is set by the physical specs of the resistor and capacitor, so no software, command, or configuration can lengthen it.

Layer 3: No Audio Ever Leaves the Sensor

Only statistics leave the sound sensor, never audio. Here's how that works:

• Once a second, the microphone captures its 87-millisecond snippet of audio and passes it to the MCU's short-term memory.
• The MCU converts that snippet into statistics for level, variability, and frequency, and keeps those numbers.
• A second later, the next snippet takes the previous one's place in memory. The previous audio is overwritten, meaning it's replaced and gone for good without ever being saved, while the statistics calculated from it are kept.
• Once a minute, the MCU combines the statistics it has gathered and sends them from the sound sensor to the monitor in a message of about 82 bytes (smaller than a single text message), and speech can't be reconstructed from those numbers.

Layer 4: Everything Is Locked at the Factory

The sound sensor's firmware, the built-in software that controls how it listens, processes, and reports sound, is locked at the factory. That means everything described in the layers above stays fixed for the life of the sensor, with no backdoor and no upgrade path for Kaiterra or anyone else.

Even if both the cloud and the monitor were compromised, and an attacker could install any firmware they wanted on the monitor, they still couldn't do anything to the sound sensor:

• They couldn't send it a command, because the link between the sensor and the monitor only runs one way.
• They couldn't replace its firmware, because the sensor has no way to receive new firmware.
• They couldn't reprogram it directly, because the access port engineers would normally use for that is locked in hardware at the factory.

On top of these layers, the sound sensor is also backed by the same platform security as every other Kaiterra device, including encrypted connections and device certificates, detailed on our security page.

Recap: Four Layers of Security to Ensure Your Privacy

Has Kaiterra's Sound Sensor Been Independently Audited?

Yes. Kaiterra gave the sensor's code and design documents to an independent security firm, Tetrel Security, which found no vulnerabilities and confirmed that the product works the way we say it does. In Tetrel's words:

An excellent design, with maximal privacy preservation.

-- Tetrel Security

Every customer who buys Kaiterra's sound sensor will also get the full design, testing method, schematic, and test results in a security white paper, so their own security team can verify every claim.

How Accurate Is a Sound Sensor That Samples 8.7% of the Time?

A sensor that listens for 87 milliseconds of every second will inevitably miss more than 90% of the sound. However, this is by design, and the missed audio makes no meaningful impact on the accuracy of the readings:

• We measured and modeled the difference. When we compared 87 milliseconds of every second against recording 100% of every second, the average difference was small enough that it won’t affect the workplace decisions the data is used for.
• Sampling a little, constantly, beats measuring everything once. Sampling 8.7% of every second, of every minute, of every day adds up to far more than 100% of a single day. An acoustic survey measures for a day or two, then the consultant goes home. Kaiterra's sensor samples every day of the year, which adds up to hundreds of hours of sound per room, per year, much more than any survey could capture.

Learn More About How Kaiterra's Sound Sensor Can Help You Deliver Acoustic Comfort

If you're in charge of a workplace, building, or portfolio, creating an acoustically comfortable space is key to your team's well-being, focus, and productivity. To learn more about how to find and fix the root causes of office noise and deliver acoustic comfort for your team without recording a single word, reach out today for a demo.