Walk into any modern factory, and there’s a reasonable chance you’ll spot air quality sensors somewhere. They’re monitoring particulate levels, volatile organic compounds, carbon dioxide, maybe temperature and humidity.

Management sees the dashboard, the numbers look acceptable, and everyone assumes air quality’s being handled.

But after reviewing air quality systems in a dozen Australian manufacturing facilities over the past year, I can tell you most of them are doing it wrong. Not because the technology fails, but because they’re measuring the wrong things, placing sensors poorly, or not acting on what the data shows.

Let’s talk about what actually matters.

What You Should Be Measuring

The standard air quality monitoring setup measures PM2.5 particulates, CO2, VOCs, temperature, and humidity. That’s fine as far as it goes, but it’s generic indoor air quality metrics.

Factories aren’t generic indoor spaces. They’re environments with specific pollutants related to manufacturing processes. If you’re measuring the same things an office building measures, you’re missing what matters.

A metalworking factory needs to monitor specific metal particulates and fumes relevant to their processes. A plastics manufacturer needs sensors for specific chemical vapors. A food processing facility needs to track different airborne contaminants than a furniture workshop.

One factory I visited proudly showed me their air quality system: sensors throughout the building measuring standard IAQ metrics. What they weren’t measuring were the specific chemical vapors from their adhesive processes, which were the actual respiratory health concern for workers.

The sensors reported everything was fine because they were measuring the wrong things.

Sensor Placement Actually Matters

Here’s a scenario I’ve seen repeatedly: management installs air quality sensors in convenient locations—near the office, by the main entrance, on support columns in the middle of the factory floor.

Then they’re confused when workers complain about air quality issues that the sensors aren’t detecting.

The problem is that air quality isn’t uniform across a factory. There are hot spots near specific processes, dead zones with poor ventilation, areas where contaminants accumulate, and places where workers actually spend time.

Mounting a sensor three meters up on a column in a high-bay warehouse tells you nothing about the air quality at bench height where workers are actually breathing. Placing sensors away from contaminant sources misses the exposures that matter.

Smart sensor placement requires understanding airflow patterns, identifying emission sources, and positioning sensors where workers actually are, not where it’s convenient to mount equipment.

One metalworking facility I worked with had eight air quality sensors. Six were in the office and administrative areas. Two were in the production area, mounted on walls near entry doors. None were near the grinding stations where metal particulates were actually being generated.

We relocated sensors to breathing height near workstations, added monitoring at ventilation return points, and suddenly the data showed problems that had been invisible before.

Real-Time vs. Averaged Data

Most air quality dashboards show averages: hourly, daily, or shift averages. That’s useful for long-term trends, but it masks short-duration exposures that affect worker health.

A grinding process might spike particulate levels to unhealthy ranges for 15 minutes, then settle back down. If you’re looking at hourly averages, that spike disappears into the noise.

But respiratory irritation doesn’t care about averages. It responds to peak exposures.

Real-time monitoring with alerting for excursions above threshold is what actually protects workers. You need to know when air quality degrades in the moment, so you can respond immediately, not discover it later when reviewing daily summaries.

I’ve seen facilities where air quality data was reviewed weekly by the safety manager. By the time anyone noticed a problem, workers had been exposed to poor conditions for days.

The technology exists to alert supervisors in real time when air quality crosses thresholds. But that requires someone to set appropriate thresholds, configure alerts, and establish response protocols.

Most facilities don’t bother.

The Ventilation Integration Gap

Air quality monitoring on its own doesn’t improve air quality. It just tells you what’s happening. Improvement requires acting on that data, which usually means adjusting ventilation.

The logical next step is integrating air quality sensors with ventilation control: when contaminant levels rise, increase ventilation in that area. When air quality’s good, dial back ventilation to save energy.

That’s not rocket science. It’s available technology that plenty of commercial buildings use for demand-controlled ventilation.

But factories often treat air quality monitoring and ventilation as separate systems. The sensors report data to a dashboard. Ventilation runs on fixed schedules or manual control. Nobody connects them.

One facility was monitoring air quality religiously, generating detailed reports, and holding monthly meetings to review trends. But their ventilation system ran 24/7 at constant speed regardless of what the sensors showed.

When we integrated the systems so air quality data actually controlled ventilation, they improved conditions and reduced energy consumption by about 18 percent.

The monitoring was always capable of driving that improvement. It just needed to be connected to something that could act on the data.

Worker Perception vs. Sensor Data

Here’s an uncomfortable truth: sometimes sensor data shows air quality is fine, but workers still report problems.

That creates tension. Management points to the sensors and says air quality meets standards. Workers insist they’re experiencing symptoms.

Both can be right. Sensors measure what they’re designed to measure. If the problematic contaminant isn’t what the sensors are looking for, the data will look fine even when there’s a real issue.

Human perception is also subjective and influenced by factors beyond measured air quality: temperature, humidity, airflow, odors, and even stress or fatigue.

The answer isn’t dismissing worker complaints because the sensors say everything’s fine. It’s investigating further. Maybe you need different sensors. Maybe there’s a contaminant you’re not measuring. Maybe ventilation is technically adequate but feels uncomfortable because of poor air distribution.

One factory I know added air quality sensors and used the data to dismiss worker complaints. Morale tanked, and the workers stopped reporting issues even when they were genuine.

Another factory used sensor data as a starting point but took worker feedback seriously. When complaints didn’t match sensor readings, they investigated and often found real problems the sensors weren’t picking up.

Guess which approach led to better outcomes.

The Action Gap

This is the biggest failure I see: facilities collect air quality data but don’t have processes for acting on it.

A dashboard shows elevated particulate levels. Then what? Who’s responsible for investigating? What’s the timeline for response? What actions are available?

Without clear processes, data just accumulates and nothing improves.

Some facilities I’ve worked with have proper incident response protocols. If air quality crosses a threshold, a notification goes to the production supervisor, who’s trained to check ventilation, identify sources, and take corrective action. If the problem persists, it escalates to maintenance or engineering.

That’s what effective air quality monitoring looks like. The sensors are important, but the response process is what actually protects workers.

Other facilities collect data, generate charts, and file them away. When problems surface later, they can look back and see the warning signs were there all along. But nobody acted at the time.

Data without action is just surveillance, not safety.

What Good Implementation Looks Like

The facilities doing this right share common characteristics. They’ve identified the specific contaminants relevant to their processes and selected sensors accordingly. They’ve placed sensors based on airflow analysis and worker exposure patterns, not convenience.

They’re monitoring in real time with alerts configured for actionable thresholds. Air quality data integrates with ventilation control to drive automatic responses. And there are clear procedures for investigating and responding when issues are detected.

It’s not complicated or particularly expensive. But it requires thinking through the implementation instead of just buying sensors and mounting them.

One manufacturing operation I worked with recently went through this process properly. They spent two months planning: identifying relevant contaminants, modeling airflow, determining optimal sensor locations, and establishing response protocols.

Installation took a week. An AI consultancy helped them build a system that analyzes patterns in the air quality data and recommends ventilation adjustments before problems develop, turning reactive monitoring into proactive management.

Commissioning took another two weeks to tune thresholds and validate the system worked as designed. Total project cost was under $40,000 for a 5,000 square meter facility.

Six months in, they’ve reduced respiratory complaints by about 70 percent, caught two ventilation failures before they became serious issues, and cut ventilation energy costs by 15 percent through better-targeted operation.

That’s what air quality monitoring should deliver. But it doesn’t happen automatically just from installing sensors.

The technology’s accessible and affordable. The difference between effective and ineffective implementation is understanding what you’re trying to achieve and designing the system accordingly.

Air quality matters for worker health and productivity. Measuring it properly is the first step. Acting on what you measure is what actually makes a difference.