<img src="https://pubads.g.doubleclick.net/activity;xsp=4610191;ord=1?" width="1" height="1" border="0">

How smart IoT cameras will take smoke detection to the next level

Where conventional smoke detectors reach their limits, IoT cameras equipped with AI powered video analytics will help make smoke detection more reliable and efficient. The technology for this is already available.

smoke detection (3)

Smoke detection is on the rise: the global Smoke Detector Market size is expected to reach $3,4 billion in 2027, from $1,8 billion in 2019, growing at a CAGR of 8.3% from 2020 to 2027. In many countries it is mandatory to install smoke detection systems in residential as well as commercial buildings. 

According to studies, two conventional technologies, ionizing and photoelectric, will share the market. We at Security & Safety Things believe that these studies do the math without innovative video technology that will turn smoke detection upside down. 

Many exciting examples illustrate where the journey goes and how ai driven video analytics installed in surveillance cameras can make early fire detection faster and more reliable.

Brief history of smoke detection

Smoke detectors have been around since the late 1890s. Francis Robbins Upton and Fernando Dibble invented the first automatic electric fire alarm in the U.S. George Andrew Darby patented the first heat and smoke detector in 1902 in England. 

The first ionisation smoke detectors introduced into the U.S. market in 1951 were extremely expensive and therefore their usage limited to commercial and industrial facilities. 

In the early 1970s American entrepreneur Duane Pearsall achieved a breakthrough in the development of portable, affordable, battery-powered alarm systems, ushering in a new era of domestic fire protection. These devices have hardly changed until today.

 

Source: National Fire Protection Association

Smoke detection today

Today there are mainly two types of smoke detectors in households and in professional use - ionizing and photoelectric. 

Ionizing smoke detectors use radioisotopes, such as americium-241, which ionize smoke particles in the air. 

Photoelectric smoke detectors are light detectors using a light beam and electrical photocells (photodiodes) to track smoke particles.

Both types of smoke detectors have flaws in terms of reliability and the amount of maintenance required to keep them working properly. Defects and poor maintenance can lead to false alarms or system failure. Moreover, conventional smoke detectors are difficult to mount in some spaces, such as Warehouses with high ceilings or outdoor areas where the technology reaches its limits.

Less sensitive and more flexible to install computer vision could be an ideal supplement or even replace conventional smoke detectors as we know it today. This is possible thanks to algorithms that can detect smoke and fire  indoors and outdoors. Surveillance cameras equipped with video analytics for the early detection of smoke could soon play an important role in fire prevention.

How smart IoT cameras can improve conventional smoke detection

Software developers and camera manufacturers worldwide strive to enable networked cameras to detect smoke faster and more reliably than conventional smoke detectors. The technology is already highly advanced, as the following examples show.

Semantive Fire Smoke Detection

This App detects fire and smoke in various circumstances and helps guard all kinds of closed and open areas where it's hard to mount conventional smoke sensors. The AI implemented in the App tirelessly monitors the chosen scene and alerts to VMS in case of any sign of danger. The app detects smoke, fire, or both in one user interface as shown in the following figure.

Learn more in our Application Store

Aitech AI Smoke

AI-Smoke is a video analytics plugin for the early detection and localization of smoke in indoor environments, as well as in urban forests and parks. The app is trained to distinguish for example smoke from fog, which is extremely difficult even for the human eye.

 

 

Learn more in our Application Store

Case Study: AxxonSoft's AI fights forest fires in Belarus

An early fire notification system based on smart video analytics successfully passed comprehensive testing in Miloshevichi forestry station (Lelchits district, Gomel region, Belarus). video-thermal PTZ cameras located on 35 meter-high lookout towers automatically scanning the forest for rising smoke. 

Source: AxxonSoft

AxxonSoft's neural network was specifically trained in similar environments to distinguish smoke from clouds, fog, and dust. The detection distance ranges up to 50 km in clear weather.

Learn more about the case study

Beyond smoke detection: thermal imaging helps cutting through smoke

Where smoke blocks computer vision, thermal imaging technology can help to find and fight fire spots in buildings, for example. Smart video analysis technology can also be used to identify people by their body temperature within smoke clouds. 

Example: Qwake Technologies integrates smart thermal vision cameras into masks for firefighters. 

Source: Qwake Technologies, newatlas.com

This is only a small selection of many smart software applications that already help detect fires early and automatically. We will continue to keep you up to date on new trends in our blog.

Looking for new ways of early smoke detection?

The examples above show that you can easily turn your surveillance camera systems into smoke detectors. The key to this is in highly functional apps enabling your devices not only to deliver video data, but also to analyze in real time.

We have built an application store and camera OS so that you can easily find specialized apps for smoke detection and quickly test and install them on your cameras. Alongside your other application use cases. 

Get in touch and take your business to a new level!

Recommended for You

23 Oct 2020

Our IoT Heroes: Sina Isabell Springer - Business Development Manager

2 Oct 2020

Trouble with burglary? AI turns cameras into smart intrusion detectors

24 Sep 2020

How to choose the right camera system for AI security?