Thermal monitoring: A 24/7 fire safety tool

AVIAN IoT CoFounder and SFO, Drew Hanover, explains how advances in AI, bispectrum imaging and connected monitoring are changing the role of thermal cameras in high-risk industrial environments.

Why is continuous thermal monitoring becoming increasingly important for industrial fire prevention today?

Continuous thermal monitoring is becoming increasingly important for many different reasons.

• Insurers are encouraging their clients to invest in advanced technologies for risk management
• Labour market challenges across the manufacturing space have led to fewer people, managing more assets
• Downtime has become increasingly expensive, avoiding downtime can translate to millions of dollars in annual savings for some manufacturers

Beyond this, we have a perpetual threat of fires in industrial environments.

When a fire happens, your customers go down the road to the next supplier – they don’t wait for you to get back on your feet. It might cost you 10-20x to replace what you built 20 years ago. Some clients have told us it’s at least an 18-month waiting time for them to replace some of their machines. They limp along with old ones, which leads to higher breakdown frequency and increased risk.

Why now? Historically, continuous monitoring flooded operators with too much unmanageable data. When we founded AVIAN, we realised the missing link wasn’t the cameras, it was interpretation and the ability to leverage the camera’s data to make decisions. By leveraging machine learning and neural networks, we can now automatically identify subtle trend shifts and diagnose critical thermal issues continuously, preventing catastrophic fires before they start.

Lastly, the cost of continuous thermal monitoring has come down dramatically. We’ve positioned ourselves to offer reliable cameras built for industrial environments at a lower price point than has been traditionally available. The days of paying $20k for a single camera are over.

How can AI-powered thermal monitoring distinguish between normal operating temperatures and genuine early signs of fire risk?

One main reason continuous thermal imaging has been held back is the inability to perceive risk directly. A hot temperature reading from a thermal camera does not necessarily imply risk.

A perfect example is in pallet yards. We’ve got customers with forklifts driving around the site all day with exhaust and engine heat signatures routinely above 300°F. If your off the shelf thermal camera only recognises temperature thresholds, spikes, etc, it’s likely to send you an alarm.

Our cameras are bi-spectrum, meaning they see both infrared and visible light using two different sensors. When a hotspot is detected, we crossreference it with the visible light image using our neural networks.

Is smoke present? Is it a forklift? Is a flame visible? Is a person welding or grinding?

By training our algorithms to understand what is creating the heat, we’ve reduced our customers’ false alarm rates by over 90% compared to traditional thermal cameras. Furthermore, because industrial environments evolve, we continuously optimise these detection models and push updates over-the-air, ensuring the system gets smarter over time.

Which industries are seeing the greatest benefit from continuous thermal imaging, and what common risks do organisations underestimate?

Our largest client bases are in the wood products and waste recycling sectors. They face very different problems. In wood manufacturing sites, heat is perpetually present due to how lumber, pallets, etc are manufactured. Hot bearings, conveyors, and saws all generate heat. Along with that comes major dust exposure. It’s a perfect storm for fire.

In waste recycling, we see a different type of problem. Batteries are everywhere nowadays, and the lithium inside these cells is prone to thermal runaway and spontaneous combustion. Typically, front end loaders drive over a battery – in these cases you must react immediately otherwise it’s going to be a really bad day.

The two most frequently underestimated risks across industrial facilities are hot work (welding, grinding) and dust accumulation. They are underestimated because traditional safety tools—like smoke detectors or manual fire watches— are reactive, only alerting you after a spark has already found a fuel source.

We built our platform to turn these passive hazards into actively managed safety metrics. Our algorithms automatically classify events like welding and grinding in real-time, effectively serving as a 24/7 digital hotwork monitor. Furthermore, we’ve developed tools that help facilities manage dust safety by continuously quantifying the effectiveness of their housekeeping programmes over time, giving management actionable data before a risk escalates. It also acts as an audit log where there is a record of every thermal event detected by the AVIAN system.

As insurers become more cautious about covering high-risk facilities, do you believe continuous thermal monitoring could become a key requirement?

We already have clients that receive premium reductions after installing AVIAN. It’s dependent on the carrier, but more of them recognise the value – largely because we have so much data to show them.

Actuarial models are changing. Insurance carriers are no longer content with annual checklists; they want continuous, verifiable risk mitigation. Because we can provide risk engineers with historical, realworld data proving our system’s efficacy, we are seeing a massive shift in how underwriters view our clients.

We have enabled several of our clients to re-gain insurance after being dropped because the risk engineers saw the steps the clients were taking to reduce their risk with AVIAN. In many cases, the insurance savings paid for the system outright, and then they had 24/7 fire safety and predictive maintenance for free.

How do you see thermal imaging evolving over the next five years?

Over the next five years, thermal imaging is moving away from being a piece of hardware you buy and toward the software that manages the data.

Regulatory requirements are tightening; facilities can no longer rely on a technician showing up once a quarter with a handheld camera just to check a compliance box.

The industry is shifting in a couple of ways:

Continuous vs. handheld inspection: These solutions are complementary, not mutually exclusive. Continuous monitoring acts like your 24/7 radar, catching a hotspot the second it develops. Handheld tools are what your maintenance team grabs to investigate and fix the physical machine once the radar alerts them.

Beyond fire safety: Thermal data shouldn’t live in its own isolated silo. At AVIAN, our focus is pushing our software to do more heavy lifting. Over the next few years, that same thermal feed will live in a central dashboard to handle predictive maintenance, track hot work and monitor dust safety all at once.

The future of this tech isn’t about manufacturing a slightly better camera. It’s about building smarter software that actually tells you what the data means so you can act before a disaster starts.